I've seen a rumor going around that OpenAI hasn't had a successful pre-training run since mid 2024. This seemed insane to me but if you give ChatGPT 5.1 a query about current events and instruct it not to use the internet it will tell you its knowledge cutoff is June 2024. Not sure if maybe that's just the smaller model or what. But I don't think it's a good sign to get that from any frontier model today, that's 18 months ago.
SemiAnalysis said it last week and AFAIK it wasn't denied.<p><a href="https://newsletter.semianalysis.com/p/tpuv7-google-takes-a-swing-at-the#:~:text=OpenAI%E2%80%99s%20leading%20researchers%20have%20not%20completed%20a%20successful%20full%2Dscale%20pre%2Dtraining%20run%20that%20was%20broadly%20deployed%20for%20a%20new%20frontier%20model%20since%20GPT%2D4o%20in%20May%202024%2C%20highlighting%20the%20significant%20technical%20hurdle%20that%20Google%E2%80%99s%20TPU%20fleet%20has%20managed%20to%20overcome." rel="nofollow">https://newsletter.semianalysis.com/p/tpuv7-google-takes-a-s...</a>
The SemiAnalysis article that you linked to stated:<p>"OpenAI’s leading researchers have not completed a successful full-scale pre-training run that was broadly deployed for a new frontier model since GPT-4o in May 2024, highlighting the significant technical hurdle that Google’s TPU fleet has managed to overcome."<p>Given the overall quality of the article, that is an uncharacteristically convoluted sentence. At the risk of stating the obvious, "that was broadly deployed" (or not) is contingent on many factors, most of which are not of the GPU vs. TPU technical variety.
My reading in between the lines is OpenAI's "GPT-5" is really a GPT-4 generation model. And this is aligned with it being unimpressive. Not the promised leap forward Altman promised.
This is misleading. They had 4.5 which was a new scaled up training run. It was a huge model and only served to pro users, but the biggest models are always used as teacher models for smaller models. Thats how you do distillation. It would be stupid to not use the biggest model you have in distillation and a waste since they have the weights.<p>The would have taken some time to calculate the efficiency gains of pretraining vs RL. Resumed the GPT-4.5 for whatever budget made sense and then spent the rest on RL.<p>Sure they chose to not serve the large base models anymore for cost reasons.<p>But I’d guess Google is doing the same. Gemini 2.5 samples very fast and seems way to small to be their base pre train. The efficiency gains in pertaining scale with model scale so it makes sense to train the largest model possible. But then the models end up super sparse and oversized and make little sense to serve in inference without distillation.<p>In RL the efficiency is very different because you have to inference sample the model to draw online samples. So small models start to make more sense to scale.<p>Big model => distill => RL<p>Makes the most theoretical sense for training now days for efficient spending.<p>So they already did train a big model 4.5. Not using it would have been absurd and they have a known recipe they could return scaling on if the returns were justified.
This is a really great breakdown. With TPUs seemingly more efficient and costing less overall, how does this play for Nvidia? What's to stop them from entering the TPU race with their $5 trillion valuation?
As others mentioned, 5T isn't money available to NVDA. It could leverage that to buy a TPU company in an all stock deal though.<p>The bigger issue is that entering a 'race' implies a race to the bottom.<p>I've noted this before, but one of NVDA's biggest risks is that its primary customers are also technical, also make hardware, also have money, and clearly see NVDA's margin (70% gross!!, 50%+ profit) as something they want to eliminate. Google was first to get there (not a surprise), but Meta is also working on its own hardware along with Amazon.<p>This isn't a doom post for NVDA the company, but its stock price is riding a knifes edge. Any margin or growth contraction will not be a good day for their stock or the S&P.
Making the hardware is actually the easy part. Everyone and their uncle who had some cash have tried by now: Microsoft, Meta, Tesla, Huawei, Amazon, Intel - the list goes on and on. But Nvidia is not a chip company. Huang himself said they are mostly a software company. And that is how they were able to build a gigantic moat. Because noone else has even come close on the software side. Google is the only one who has had some success on this side, because they also spent tons of money and time on software refinement by now, while all the other chips vanished into obscurity.
Are you saying that Google, Meta, Amazon, etc... can't do software? It's the bread and butter of these companies. The CUDA moat is important to hold off the likes of AMD, but hardware like TPUs for internal use or other big software makers is not a big hurdle.<p>Of course Huang will lean on the software being key because he sees the hardware competition catching up.
Genuine question: given LLMs' inexorable commoditization of software, how soon before NVDA's CUDA moat is breached too? Is CUDA somehow fundamentally different from other kinds of software or firmware?
> What's to stop them from entering the TPU race with their $5 trillion valuation?<p>Valuation isn’t available money; they'd have to raise more money in the current, probably tighter for them, investment environment to enter the TPU race, since the money they have already raised that that valuation is based on is already needed to provide runway for what they are already doing <i>without</i> putting money into the TPU race
$5 trillion valuation doesn't mean it has $5 trillion cash in pocket -- so "it depends"
Dylan Patel joined Dwarkesh recently to interview Satya Nadella: <a href="https://www.dwarkesh.com/p/satya-nadella-2" rel="nofollow">https://www.dwarkesh.com/p/satya-nadella-2</a>
That is.... actually a seriously meaty article from a blog I've never heard of. Thanks for the pointer.
Semi analysis is great, they typically do semiconductors but reporting is top notch.
Semianalysis is great, def recommend following
Dylan Patel founded Semianalysis and he has a great interview with Satya Nadella on Dwarkesh Patel's podcast.
It's not a rumor, it's confirmed by OpenAI. All "models" since 4o are actually just optimizations in prompting and a new routing engine. The actual -model- you are using with 5.1 is 4. Nothing has been pre-trained from scratch since 4o.<p>Their own press releases confirm this. They call 5 their best new "ai system", not a new model<p><a href="https://openai.com/index/introducing-gpt-5/" rel="nofollow">https://openai.com/index/introducing-gpt-5/</a>
I don't think that counts as confirmation. 4.5 we know was a new base-model. I find it very very unlikely the base model of 4 (or 4o) is in gpt5. Also 4o is a different base model from 4 right? it's multimodal etc. Pretty sure people have leaked sizes etc and I don't think it matches up.
I can believe this, Deepseek V3.2 shows that you can get close to "gpt-5" performance with a gpt-4 level base model just with sufficient post-training.
New AI system doesn't preclude new models. I thought when GPT 5 launched and users hated it the speculation was GPT 5 was a cost cutting model and the routing engine was routing to smaller, specialized dumber models that cost less on inference?<p>It certainly was much dumber than 4o on Perplexity when I tried it.
> and the routing engine was routing to smaller, specialized dumber models that cost less on inference?<p>That this was part of it was <i>stated outright</i>, except maybe that they "cost less" which was left for you to infer (sorry), in their launch announcement.<p>Paying for pro, and setting it to thinking all the time, I saw what seemed like significant improvements, but if your requests got (mis-)routed to one of the dumber models, it's not surprising if people were disappointed.<p>I think they made a big mistake in not clearly labelling the responses with which of the models responded to a given request, as it made people complain about GPT 5 in general, instead of complaining about the routing.
Well then 5.x is pretty impressive
Maybe this is just armchair bs on my part, but it seems to me that the proliferation of AI-spam and just general carpet bombing of low effort SEO fodder would make a lot of info online from the last few years totally worthless.<p>Hardly a hot take. People have theorized about the ouroboros effect for years now. But I do wonder if that’s part of the problem
Every so often I try out a GPT model for coding again, and manage to get tricked by the very sparse conversation style into thinking it's great for a couple of days (when it says nothing and then finishes producing code with a 'I did x, y and z' with no stupid 'you're absolutely' right sucking up and it works, it feels very good).<p>But I always realize it's just smoke and mirrors - the actual quality of the code and the failure modes and stuff are just so much worse than claude and gemini.
I am a novice programmer -- I have programmed for 35+ years now but I build and lose the skills moving between coder to manager to sales -- multiple times. Fresh IC since last week again :) I have coded starting with Fortran, RPG and COBOL and I have also coded Java and Scala. I know modern architecture but haven't done enough grunt work to make it work or to debug (and fix) a complex problem. Needless to say sometimes my eyes glaze over the code.<p>And I write some code for my personal enjoyment, and I gave it to Claude 6-8 months back for improvement, it gave me a massive change log and it was quite risky so abandoned it.<p>I tried this again with Gemini last week, I was more prepared and asked it to improve class by class, and for whatever reasons I got better answers -- changed code, with explanations, and when I asked it to split the refactor in smaller steps, it did so. Was a joy working on this over the thanksgiving holidays. It could break the changes in small pieces, talk through them as I evolved concepts learned previously, took my feedback and prioritization, and also gave me nuanced explanation of the business objectives I was trying to achieve.<p>This is not to downplay claude, that is just the sequence of events narration. So while it may or may not work well for experienced programmers, it is such a helpful tool for people who know the domain or the concepts (or both) and struggle with details, since the tool can iron out a lot of details for you.<p>My goal now is to have another project for winter holidays and then think through 4-6 hour AI assisted refactors over the weekends. Do note that this is a project of personal interest so not spending weekends for the big man.
> I was more prepared and asked it to improve class by class, and for whatever reasons I got better answers<p>There is a learning curve with all of the LLM tools. It's basically required for everyone to go through the trough of disillusionment when you realize that the vibecoding magic isn't quite real in the way the influencers talk about it.<p>You still have to be involved in the process, steer it in the right direction, and review the output. Rejecting a lot of output and re-prompting is normal. From reading comments I think it's common for new users to expect perfection and reject the tools when it's not vibecoding the app for them autonomously. To be fair, that's what the hype influencers promised, but it's not real.<p>If you use it as an extension of yourself that can type and search faster, while also acknowledging that mistakes are common and you need to be on top of it, there is some interesting value for some tasks.
It really depends on what you're building. As an experiment, I started having Claude Code build a real-time strategy game a bit over a week ago, and it's done an amazing job, with me writing no code whatsoever. It's an area with lots of tutorials for code structure etc., and I'm guessing that helps. And so while I've had to <i>read</i> the code and tell it to refactor things, it has managed to do a good job of it with just relatively high level prodding, and produced a well-architected engine with traits based agents for the NPCs and a <i>lot</i> of well-functioning game mechanics. It started as an experiment, but now I'm seriously toying with building an actual (but small) game with it just to see how far it can get.<p>In other areas, it is as you say and you need to be on top of it constantly.<p>You're absolutely right re: the learning curve, and you're much more likely to hit an area where you need to be on top of it than one that it can do autonomously, at least without a lot of scaffolding in the form of sub-agents, and rules to follow, and agent loops with reviews etc., which takes a lot of time to build up, and often include a lot of things specific to what you want to achieve. Sorting through how much effort is worth it for those things for a given project will take time to establish.
For me the learning curve was learning to choose what is worth asking to Claude. After 3 months on it, I can reap the benefit: Claude produces the code I want right 80% of the time. I usually ask it: to create new functions from scratch (it truly shines at understanding the context of these functions by reusing other parts of the code I wrote), refactor code, create little tools (for example a chart viewer).
I appreciate this narrative; relatable to me in how I have experienced and watched others around me experience the last few years. It's as if we're all kinda-sorta following a similar "Dunning–Kruger effect" curve at the same time. It feels similar to growing up mucking around with a ppp connection and Netscape in some regards. I'll stretch it: "multimodal", meet your distant analog "hypermedia".
My problem with Gemini is how token hungry it is. It does a good job but it ends up being more expensive than any other model because it's so yappy. It sits there and argues with itself and outputs the whole movie.
Breaking down requirements, functionality and changes into smaller chunks is going to give you better results with most of the tools. If it can complete smaller tasks in the context window, the quality will likely hold up.
My go to has been to develop task documents with multiple pieces of functionality and sub tasks. Build one piece of functionality at a time. Commit, clear context and start the next piece of functionality.
If something goes off the rails, back up to the commit, fix and rebase future changes or abandon and branch.<p>That’s if I want quality. If I just want to prototype and don’t care, I’ll let it go. See what I like, don’t like and start over as detailed above.
Interesting. From my experience, Claude is much better at stuff involving frontend design somehow compared to other models (GPT is pretty bad). Gemini is also good but often the "thinking" mode just adds stuff to my code that I did not ask it to add or modifies stuff to make it "better". It likes to 1 up on the objective a lot which is not great when you're just looking for it to do what you precisely asked it and nothing else.
I have never considered trying to apply Claude/Gemini/etc. to Fortran or COBOL. That would be interesting.
You can actually use Claude Code (and presumably the other tools) on non-code projects, too. If you launch claude code in a directory of files you want to work on, like CSVs or other data, you can ask it to do planning and analysis tasks, editing, and other things. It's fun to experiment with, though for obvious reasons I prefer to operate on a copy of the data I'm using rather than let Claude Code go wild.
I use Claude Code for "everything", and have just committing most things into git as a fallback.<p>It's great to then just have it write scripts, and then write skills to use those scripts.<p>A lot of my report writing etc. now involve setting up a git repo, and use Claude to do things like process the call transcripts from discovery calls and turn them into initial outlines and questions that needs followup, and tasks lists, and write scripts to do necessary analysis etc., so I can focus on the higher level stuff.
Side note from someone who just used Claude Code today for the first time: Claude Code is a TUI, so you can run it in any folder/with any IDE and it plays along nicely. I thought it was just another vscode clone, so I was pleasantly surprised that it didn't try to take over my entire workflow.
It's even better: It's a TUI if you launch it without options, but you can embed it in scripts too - the "-p" option takes a prompt, in which case it will return the answer, and you can also provide a conversation ID to continue a conversation, and give it options to return the response as JSON, or stream it.<p>Many of the command line agent tools support similar options.
I was just giving my history :) but yes I am sure this could actually get us out of the COBOL lock-in which requires 70 years old programmers to continue working.<p>The last article I could find on this is from 2020 though: <a href="https://www.cnbc.com/2020/04/06/new-jersey-seeks-cobol-programmers-to-fix-unemployment-system.html" rel="nofollow">https://www.cnbc.com/2020/04/06/new-jersey-seeks-cobol-progr...</a>
I'm starting with Claude at work but did have an okay experience with OpenAi so far. For clearly delimited tasks it does produce working code more often than not. I've seen some improvement on their side compared to say, last year. For something more complex and not clearly defined in advance, yes, it does produce plausible garbage and it goes off the rails a lot. I was migrating a project and asked ChatGPT to analyze the original code base and produce a migration plan. The result seemed good and encouraging because I didn't know much about that project at that time. But I ended up taking a different route and when I finished the migration (with bits of help from ChatGPT) I looked at the original migration plan out of curiosity since I had become more familiar with the project by now. And the migration plan was an absolutely useless and senseless hallucination.
On the contrary, I cannot use the top Gemini and Claude models because their outputs are so out place and hard to integrate with my code bases. The GPT 5 models integrate with my code base's existing patterns seamlessly.
I've been getting great results from Codex. Can be a bit slow, but gets there. Writes good Rust, powers through integration test generation.<p>So (again) we are just sharing anecdata
NME at all - 5.1 codex has been the best by far.
By my tests (<a href="https://github.com/7mind/jopa" rel="nofollow">https://github.com/7mind/jopa</a>) Gemini 3 is somewhat better than Claude with Opus 4.5. Both obliterate Codex with 5.1
Codex is super cheap though even with the cheapest GPT subscription you get lots of tokens. I use 4.5 opus at work and codex at home tbh the differences are not that big if you know what you are doing.
What's - roughly - your monthly spend when using ppt models? I only use fixed priced copilot, and my napkin maths says I'd be spending something crazy like $200/mo if I went ppt on the more expensive models.
They have subscriptions too (at least Claude and ChatGPT/Codex; I don't use Gemini much). It's far cheaper to use the subscriptions first and then switch to paying per token beyond that.
Something around 500 euros.
How can you stand the excruciating slowness? Claude Code is running circles around codex. The most mundane tasks make it think for a minute before doing anything.
I use it on medium reasoning and it's decently quick. I only switch to gpt-5.1-codex-max xhigh for the most annoying problems.
By learning to parallelize my work. This also solved my problem with slow Xcode builds.
Well you can’t edit files while Xcode is building or the compiler will throw up, so I‘m wondering what you mean here. You can’t even run swift test in 2 agents at the same time, because swift serializes access for some reason.<p>Whenever I have more than 1 agent run Swift tests in a loop to fix things, and another one to build something, the latter will disturb the former and I need to cancel.<p>And then there’s a lot of work that can’t be parallelized, like complex git rebases - well you can do other things in a worktree, but good luck merging that after you‘ve changed everything in the repo. Codex is really really bad at git.
I use the web ui, easy to parallelize stuff to 90% done. manually finish the last 10% and a quick test
Yes these are horrible pain points. I can only hope Apple improves this stuff if it's true that they're adding MCP support throughout the OS which should require better multi-agent handling<p>You can use worktrees to have multiple copies building or testing at once<p>I'm a solo dev so I rarely use some git features like rebase. I work out of trunk only without branches (if I need a branch, I use a feature flag). So I can't help with that<p>What I did is build an Xcode MCP server that controls Xcode via AppleScript and the simulator via accessibility & idb. For running, it gives locks to the agent that the agent releases once it's done via another command (or by pattern matching on logs output or scripting via JS criteria for ending the lock "atomically" without requiring a follow-up command, for more typical use). For testing, it serializes the requests into a queue and blocks the MCP response.<p>This works well for me because I care more about autonomous parallelization than I do eliminating waiting states, as long as I myself am not ever waiting. (This is all very interesting to me as a former DevOps/Continuous Deployment specialist - dramatically different practices around optimizing delivery these days...)<p>Once I get this tool working better I will productize it. It runs fully inside the macOS sandbox so I will deploy it to the Mac App Store and have an iOS companion for monitoring & managing it that syncs via iCloud and TailScale (no server on my end, more privacy friendly). If this sounds useful to you please let me know!<p>In addition to this, I also just work on ~3 projects at the same time and rotate through them by having about 20 iTerm2 tabs open where I use the titles of each tab (cmd-i to update) as the task title for my sake.<p>I've also started building more with SwiftWASM (with SQLite WASM, and I am working on porting SQLiteData to WASM too so I can have a unified data layer that has iCloud sync on Apple platforms) and web deployment for some of my apps features so that I can iterate more quickly and reuse the work in the apps.
Yes, that makes sense to me. I cannot really put builds in a queue because I have very fine-grained updates that I tell my agents so they do need the direct feedback to check what they have just done actually works, or they will interfere with each other’s work.<p>I do strive to use Mac OS targets because those are easier to deal with than a simulator, especially when you use Bluetooth stuff and you get direct access to log files and SQLite files.<p>Solo devs have it way easier in this new world because there’s no strict rules to follow. Whatever goes, goes, I guess.
How can I get in touch?
[dead]
Same experience here. The more commonly known the stuff it regurgitates is, the fewer errors. But if you venture into RF electronics or embedded land, beware of it turning into a master of bs.<p>Which makes sense for something that isn’t AI but LLM.
Can you give some concrete example of programming problem task GPT fails to solve?<p>Interested, because I’ve been getting pretty good results with different tasks using the Codex.
Try to ask it to write some GLSL shaders. Just describe what you want to see and then try to run the shaders it outputs. It can output a UV-map or the simple gradient right, but when it comes to shaders a bit more complex it most of the time will not compile or run properly, sometimes mix GLSL versions, sometimes just straight make up things which don't work or output what you want.
I posted this example before but academic papers on algorithms often have pseudo code but no actual code.<p>I thought it would be handy to use AI to make the code from the paper so a few months ago I tried to use Claude (not GPT, because I only have access to Claude) to recreate C++ code to implement the algorithms in this paper as practice for me in LLM use and it didn’t go well.<p><a href="https://users.cs.duke.edu/~reif/paper/chen/graph/graph.pdf" rel="nofollow">https://users.cs.duke.edu/~reif/paper/chen/graph/graph.pdf</a>
Completely failed for me running the code it changed in a docker container i keep running. Claude did it flawlessly.
It absolutely rocks at code reviews but ir‘s terrible in comparison generating code
You're absolutely right!<p>Somehow it doesn't get on my nerves (unlike Gemini with "Of course").
I find for difficult questions math and design questions GPT5 tends to produce better answers than Claude and Gemini.
At this point you are now <i>forced</i> to use the "AI"s as code search tools--and it annoys me to no end.<p>The problem is that the "AI"s can cough up code examples based upon proprietary codebases that you, as an individual, have no access to. That creates a significant quality differential between coders who only use publicly available search (Google, Github, etc.) vs those who use "AI" systems.
OpenAI is the only SOTA model provider that doesn't have a cutoff date in the current year. That why it preforms bad at writing code for any new libraries or libraries that have had significant updates like Svelte.
OpenAI is in the "don't look behind the curtain" stage with both their technology and finances.
I recall reading that Google had similar 'delay' issues when crawling the web in 2000 and early 2001, but they managed to survive. That said, OpenAI seems much less differentiated (now) than Google was back then, so this may be a much riskier situation.
Google didn't raise at a $500 billion valuation.<p>The 25x revenue multiple wouldn't be so bad if they weren't burning so much cash on R&D and if they actually had a moat.<p>Google caught up quick, the Chinese are spinning up open source models left and right, and the world really just isn't ready to adopt AI everywhere yet. We're in the premature/awkward phase.<p>They're just too early, and the AGI is just too far away.<p>Doesn't look like their "advertising" idea to increase revenue is working, either.
The differentiation should be open source, nonprofit, and ethical.<p>As a shady for-profit, there is none. That's the problem with this particular fraud.
Yes, the story was something like Google hadn’t rebuilt their index for something like 8 months if I recall correctly
I noticed this recently when I asked it whether I should play Indiana Jones on my PS5 or PC with a 9070 XT. It assumed I had made a typo until I clarified, then it went off to the internet and came back telling me what a sick rig I have.
I asked ChatGPT 5.1 to help me solve a silly installation issue with the codex command line tool (I’m not an npm user and the recommended installation method is some kludge using npm), and ChatGPT told me, with a straight face, that codex was discontinued and that I must have meant the “openai” command.
Just a minor correction, but I think it's important because some comments here seem to be giving bad information, but on OpenAI's model site it says that the knowledge cutoff for gpt-5 is Sept 30, 2024, <a href="https://platform.openai.com/docs/models/compare" rel="nofollow">https://platform.openai.com/docs/models/compare</a>, which is later than the June 01, 2024 date of GPT-4.1.<p>Now I don't know if this means that OpenAI was able to add that 3 months of data to earlier models by tuning or if it was a "from scratch" pre-training run, but it has to be a substantial difference in the models.
I doubt it's that important that their dataset of current events is up to date. At this stage, I believe private and synthetic data comprises a large fraction of pretraining. Web search substitutes for current event pretraining.
Funny, had it tell me the same thing twice yesterday and that was _with_ thinking + search enabled on the request (it apparently refused to carry out the search, which it does once in every blue moon).<p>I didn't make this connection that the training data is that old, but that would indeed augur poorly.
What is a pre-training run?
Pre-training is just training, it got the name because most models have a post-training stage so to differentiate people call it pre-training.<p>Pre-training: You train on a vast amount of data, as varied and high quality as possible, this will determine the distribution the model can operate with, so LLMs are usually trained on a curated dataset of the whole internet, the output of the pre-training is usually called the base model.<p>Post-training: You narrow down the task by training on the specific model needs you want. You can do this through several ways:<p>- Supervised Finetuning (SFT): Training on a strict high quality dataset of the task you want. For example if you wanted a summarization model, you'd finetune the model on high quality text->summary pairs and the model would be able to summarize much better than the base model.<p>- Reinforcement Learning (RL): You train a separate model that ranks outputs, then use it to rate the output of the model, then use that data to train the model.<p>- Direct Preference Optimizaton (DPO): You have pairs of good/bad generations and use them to align the model towards/away the kinds of responses you want.<p>Post-training is what makes the models able to be easily used, the most common is instruction tuning that teaches to model to talk in turns, but post-training can be used for anything. E.g. if you want a translation model that always translates a certain way, or a model that knows how to use tools, etc. you'd achieve all that through post-training. Post-training is where most of the secret sauce in current models is nowadays.
If pre-training is just training, then how on earth can OpenAI not have "a successful pre-training run"? The word successful indicates that they tried, but failed.<p>It might be me misunderstanding how this works, but I assumed that the training phase was fairly reproducible. You might get different results on each run, do to changes in the input, but not massively so. If OpenAI can't continuously and reliably train new models, then they are even more overvalued that I previously assumed.
Want to also add that the model doesn’t know how to respond in a user-> assistant style conversation after it’s pretraining, and it’s a pure text predictor (look at the open source base models)<p>There’s also what is being called mid-training where the model is trained on high(er) quality traces and acts as a bridge between pre and post training
If you've an hour to spare this Karpathy video is good at explaining how it all works <a href="https://youtu.be/7xTGNNLPyMI" rel="nofollow">https://youtu.be/7xTGNNLPyMI</a>
The first step in building a large language model. That's when the model is initiated and trained on a huge dataset to learn patterns and whatnot. The "P" in "GPT" stands for "pre-trained."
That’s where they take their big pile of data and train the model to do next-token-prediction.
That would explain why it’s so bad with new Swift features and more recent ast-grep rules.
Usually current events get taught through mid-training, so even with old pre-training current events still could be added
Don’t forget SemiAnalysis’s founder Dylan Patel is supposedly roommates with Anthropics RL tech lead Sholto..
The fundamental problem with bubbles like this, is that you get people like this who are able to take advantage of the The Gell-Mann amnesia effect, except the details that they’re wrong about are so niche that there’s a vanishingly small group of people who are qualified to call them out on it, and there’s simultaneously so much more attention on what they say because investors and speculators are so desperate and anxious for new information.<p>I followed him on Twitter. He said some very interesting things, I thought. Then he started talking about the niche of ML/AI I work near, and he was completely wrong about it. I became enlightened.
It has no idea what it's own knowledge cutoff is.
Any data after that is contaminated with vast amount of AI slop. Is anyone training on anything newer..?
Interesting enough today I found that <a href="https://huggingface.co/openai/gpt-oss-20b" rel="nofollow">https://huggingface.co/openai/gpt-oss-20b</a> when downloaded an running locally and asked:<p>====<p>when have you been built?<p>I’m a GPT‑4‑based model that OpenAI released on March 14 2023.
The underlying training data goes up to early 2023, and the model was trained in the months leading up to that release. If you’re asking about the ChatGPT product that ships the model to users, it went live in November 2022 and has since received updates (GPT‑3.5, GPT‑4, etc.) that keep it current.<p>====<p>But when supposedly running it from <a href="https://huggingface.co/chat/models/openai/gpt-oss-20b" rel="nofollow">https://huggingface.co/chat/models/openai/gpt-oss-20b</a>:<p>====<p>when have you been built?<p>I’m a language model created by OpenAI. The current generation (GPT‑4) that powers this chat was first released in March 2023 and has been updated and fine‑tuned up through the end of 2024. My training data runs up to the beginning of June 2025, so I’m built on knowledge available up to that point.<p>====<p>And that makes me thinking that although <a href="https://huggingface.co/chat" rel="nofollow">https://huggingface.co/chat</a> claims to be using the models available to public at <a href="https://huggingface.co" rel="nofollow">https://huggingface.co</a> , it doesn't seems to be true and I raised this question here <a href="https://huggingface.co/ggml-org/gpt-oss-20b-GGUF/discussions/4" rel="nofollow">https://huggingface.co/ggml-org/gpt-oss-20b-GGUF/discussions...</a> , <a href="https://github.com/huggingface/inference-playground/issues/102" rel="nofollow">https://github.com/huggingface/inference-playground/issues/1...</a> and <a href="https://github.com/ggml-org/llama.cpp/discussions/15396#discussioncomment-15136920" rel="nofollow">https://github.com/ggml-org/llama.cpp/discussions/15396#disc...</a> .
OpenAI is basically just Netscape at this point. An innovative product with no means of significant revenue generation.<p>One one side it's up against large competitors with an already established user base and product line that can simply bundle their AI offerings into those products. Google will do just what Microsoft did with Internet Explorer and bundle Gemini in for 'Free' with their already other profitable products and established ad-funded revenue streams.<p>At the same time, Deepseek/Qwen, etc. are open sourcing stuff to undercut them on the other side. It's a classic squeeze on their already fairly dubious business model.
> with no means of significant revenue generation.<p>OpenAI will top $20 billion in ARR this year, which certainly seems like significant revenue generation. [1]<p>[1] <a href="https://www.cnbc.com/2025/11/06/sam-altman-says-openai-will-top-20-billion-annual-revenue-this-year.html" rel="nofollow">https://www.cnbc.com/2025/11/06/sam-altman-says-openai-will-...</a>
I can generate $20 billion in ARR this year too! I just need you to give me $100 billion and allow me to sell each of your dollars for 0.2 dollars.
It's a fun trope to repeat but that's not what OpenAI is doing. I get a ton of value from ChatGPT and Codex from my subscription. As long as the inference is not done at a lost this analogy doesn't hold. They're not paying me to use it. They are generating output that is very valuable to me. Much more than my subscription cost.<p>I've been able to help setup cross app automation for my partner's business, remodel my house, plan a trip of Japan and assist with the cultural barrier, vibe code apps, technical support and so much more.
To be fair, I would get a ton of value out of someone selling dollars for 20 cents apiece.<p>But ya, OAI is clearly making a ton of revenue. That doesn't mean it's a good business, though. Giving them a 20 year horizon, shareholders will be very upset unless the firm can deliver about a trillion in <i>profit</i>, not revenue, to justify the 100B (so far) in investment, and that would barely beat the long term s&p 500 average return.<p>But Altman himself has said he'll need much more investment in the coming years. And even if OAI became profitable by jacking up prices and flooding gpt with ads, the underlying technology is so commodified, they'd never be able to achieve a high margin, assuming they can turn a profit at all.
People did say the same thing about Youtube, which was unprofitable and extremely expensive to run in the early years. I remember thinking everyone would leave once ads were added.<p>At youtube's ad income rate (~$13/year), the current (but growing) ~800 million chatgpt users would add ~$10 billion. At facebook's rate (~$40-50/year) $32-40 billion. Potentially, an assistant would be more integrated into your life than either of those two.<p>The "audience retention" is the key question, not the profitability if they maintain their current audience. I've been surprised how many non-technical people I know don't want to try other models. "ChatGPT knows me".
The network effects aren't the same. All the viewers watch youtube because it has all the content, and all the creators post on youtube because it has all the viewers.<p>How can a model achieve this kind of stickiness? By "knowing you"? I don't think that's the same at all. Personally, one of the reasons I prefer Claude is that it doesn't pretend to know me. I can control the context better.
the problem with the YouTube analogy is that media platforms have significant network affects that NN providers don't. OpenAI can't command a premium because every year that goes by the cost to train an equivalent model to theirs decreases.
Youtube didn't either at the time. The front page was widely seen as garbage, and everyone I knew watched videos because they were embedded or linked from external sites. "If they introduced ads, people will just switch to other video hosts, wont they?". Many of the cooler creators used Vimeo. It was the good recommendation algorithm that came later, that I think allowed an actual network effect, and I don't remember people predicting that.<p>The field is too young to know what will keep users, but there are definitely things that plausibly could create a lock-in effect. I mentioned one ("ChatGPT knows me") which could grow over time as people have shared more of themselves with ChatGPT. There's also pilots of multi-person chats, and the social elements in Sora. Some people already feel compelled to stick to the "person" they're comfortable talking to. The chance of OpenAI finding something isn't zero.
I suspect some of the downvoters hate the idea of ads, which is understandable.<p>But a lot of HN users use gmail, which has the same model. And there are plenty of paid email providers which seem far less popular (I use one). Ads didn't end up being a problem for most people provided they were kept independent of the content itself.
I'd be a little bit more nuanced:<p>I think there's something off with their plans right now: it's pretty clear at this point that they can't own the technological frontier, Google is just too close already and from a purely technological PoV they are much better suited to have the best tech in the medium term. (There's no moat and Google has way more data and compute available, and also tons of cash to burn without depending on external funding).<p>But ChatGPT is an insane <i>brand</i> and for most (free) customers I don't think model capabilities (aka “intelligence”) are that important. So if they stopped training frontier models right now and focus on driving their costs low by optimizing their inference compute budget while serving ads, they can <i>make a lot of money</i> from their user base.<p>But that would probably mean losing most of its paying customers over the long run (companies won't be buying mediocre token at a premium for long) and more importantly it would require abandoning the AGI bullshit narrative, which I'm not sure Altman is willing to do. (And even if he was, how to do that without collapsing from lack of liquidity due to investors feeling betrayed is an open question).
Being an insane brand means literally nothing if people can trivially switch to competitors, which they can.<p>There isn't even a tenth of enough money if you group together all of advertising. Like, the entire industry. Ads is a bad, bad plan that wont work. Advertising is also extremely overvalued. And even at it's overvalued price tag, it's nowhere near enough.
People could trivially switch their search engine to Bing or Yahoo, but they don't.<p>If ads are so overpriced, how big is your short position on google? Also ads are extremely inefficient in terms of conversion. Ads rendered by an intelligent, personalized system will be OOM more efficient, negating most of the "overvalue".<p>I'm not saying they should serve ads. It's a terrible strategy for other reasons.
Funny that you mention Yahoo, as in my mind they're the perfect example of what the poster above you noted: people quickly switched to Google once a better alternative to Yahoo appeared.
You know that Google literally spends billions to ensure that people don’t switch, right?<p>That’s possible because they’re immensely profitable.
It's Coca Cola vs Pepsi. Yes some might even say Pepsi has been shown to taste better, but people still buy loads of Coke.<p>Of course the tech savvy enterprises will use the best models. But the plumber down the road doesn't care whether she asks Gemini or ChatGPT about the sizing of some fittings.
> Being an insane brand means literally nothing if people can trivially switch to competitors, which they can.<p>Logically speaking, yes it is easy to switch between OAI and Gemini, or Coke and Pepsi. But brand loyalty is more about emotions (comfort, familiarity,..) rather logical reasoning.
The best way to drive inference cost down right now is to use TPUs. Either that or invest tons of additional money and manpower into silicon design like Google did, but they already have a 10 year lead there.
Altman's main interest is Altman. ChatGPT will be acquihired, most people will be let go, the brand will become a shadow of its former self, and Altman will emerge with a major payday and no obvious dent in his self-made reputation as a leading AGIthinkfluenceretc.<p>I don't think ads are that easy, because the hard part of ads isn't taking money and serving up ad slop, it's providing convincing tracking and analytics.<p>As soon as ad slop appears a lot of customers will run - not all, but enough to make monetisation problematic.
as long as the business model is:<p>- users want the best/smartest LLM<p>- the best performance for inference is found by spending more and more tokens (deep thinking)<p>- pricing is based on cost per token<p>Then the inference providers/hyperscalers will take all of the margin available to app makers (and then give it to Nvidia apparently). It is a bad business to be in, and not viable for OpenAI at their valuation.
> But ChatGPT is an insane brand<p>I mean, so was netscape.
This. Netscape was THE browser in the early phases of the Internet. Then Microsoft just packaged IE into Windows and it was game over. The brand means nothing long term. If Google broadly incorporates Gemini into all the Google-owned things everyone already has then it’s game over for OpenAI.<p>The mass commoditization of the tech is rapidly driving AI to be a feature, not a product. And Google is very strongly positioned to take advantage of that. Microsoft too, and of course they have a relationship with OpenAI but that’s fraying.
To be completely fair the later versions of Netscape were increasingly giant bloated piles of crap while IE slowly caught up and surpassed in terms of speed and features. The first versions IE were only good for downloading Netscape.<p>Netscape, to a large degree, killed itself.<p>Not to say IE turned into anything good though. But it did have its hayday.
Maybe, I was too young to remember that.
All of which you will be able to do with your bundled assistant in the not-to-distant future.<p>OpenAI is a basket case:<p>- Too expensive and inconvenient to compete with commoditized, bundled assistants (from Google/ Microsoft/Apple)<p>- Too closed to compete with cheap, customizable open-source models<p>- Too dependent on partners<p>- Too late to establish its own platform lock-in<p>It echoes what happened to:<p>- Netscape (squeezed by Microsoft bundling + open protocols)<p>- BlackBerry (squeezed by Apple ecosystem + open Android OS)<p>- Dropbox (squeezed by iCloud, Google Drive, OneDrive + open tools like rclone)<p>When you live between giants and open-source, your margin collapses from both sides.
There's no doubt you're getting a lot of value from OpenAI, I am too. And yes the subscription is a lot more value than what you pay for. That's because they're burning investor's money and it's not something that is sustainable. Once the money runs out they'll have to jack up prices and that's the moment of truth, we'll see what users are willing to pay for what. Google or another company may be able to provide all that much cheaper.
> They're not paying me to use it.<p>Of course they are.<p>> As long as the inference is not done at a loss.<p>If making money on inference alone was possible, there would be a dozen different smaller providers who'd be taking the open weights models and offering that as service. But it seems that every provider is anchored at $20/month, so you can bet that none of them can go any lower.
> If making money on inference alone was possible, there would be a dozen different smaller providers who'd be taking the open weights models and offering that as service.<p>There are! Look through the provider list for some open model on <a href="https://openrouter.ai" rel="nofollow">https://openrouter.ai</a> . For instance, DeepSeek 3.1 has a dozen providers. It would not make any sense to offer those below cost because you have neither moat nor branding.
> If making money on inference alone was possible<p>Maybe, but arguably a major reason you can't make money on inference right now is that the useful life of models is too short, so you can't amortize the development costs across much time because there is so much investment in the field that everyone is developing new models (shortening useful life in a competitive market) and everyone is simultaneously driving up the costs of inputs needed for developing models (increasing the costs that have to be amortized over the short useful life). Perversely, the AI bubble popping and resolving those issues may make profitability much easier for the survivors that have strong revenue streams.
They do make money on inference.
The open models suck. AWS hosts them for less than closed models cost but no ones uses them, because they suck.
> I get a ton of value from ChatGPT and Codex from my subscription<p>I think that’s what they’re saying. OpenAI is selling you a $1 product for $0.2<p>Tokens are too cheap right now and nobody is working on a path to dial up the cost
As a developer - ChatGPT doesn't hold a candle compared to claude for coding related tasks and under performs for arbitrary format document parsing[1]. It still has value and can handle a lot of tasks that would amaze someone in 2020 - but it is simply falling behind and spending much more doing so.<p>1. It actually under performs Claude, Gemini and even some of the Grok models for accuracy with our use case of parsing PDFs and other rather arbitrarily formatted files.
Well, don't you think you're getting a ton of value because they're selling each of their dollars for 0.2 dollars?
That the product is useful does not mean the supplier of the product has a good business; and of course, vice versa. OpenAI has a <i>terrible</i> business at the moment, and the question is, do they have a plausible path to a good one?
If the subscription cost 5x as much would you still pay and feel you are getting such a great value?
If there are no free alternatives, yes. 100 USD a month for ChatGPT seems great value
I pay $100/month for Claude Max, and I've already said it, I would go up to $500 a month and wouldn't hesitate for a second. I'd probably start to hesitate for $1,000 maybe, only cuz I know I wouldn't be able to use it enough to maximize that value. But I might still suck it up and pay for it (I don't use it enough yet to need the $200/month but if I started hitting limits faster, I would upgrade), or at that point start looking for alternatives.<p>It's worth that much to me in the time saved. But I'm a business owner, so I think the calculus might be quite different (since I can find ways to recoup those costs) from an individual, who pays out of their main income.<p>I outlined examples of how I used CC/AI a couple months ago [1]. Since then I've used it even more, to help reduce our cloud bills.<p>1: <a href="https://news.ycombinator.com/item?id=45382337">https://news.ycombinator.com/item?id=45382337</a>
I mean Claude is good for business use-cases, other than that it's completely censored cuck garbage and the CEO is worse than the pope. With Grok you can actually roleplay without it wagging its finger at you. OH MY GOSH YOU SAID BOOB!<p>Normies literally see no difference between GPT and Claude, just that Claude is much more expensive and CEO is even more of a dummie than Altman.
You are mostly missing the point. You’re saying you get value out of what OpenAI is offering you. Thats not at issue here.<p>The question is, does OpenAI get value out of the exchange?<p>You touched on it ever so briefly: “as long as inference is not done at a loss”. That is it, isn’t it? Or more generally, As long as OpenAI is making money . But they are not.<p>There’s the rub.<p>It’s not only about whether you think giving them your money is a good exchange. It needs to be a good exchange for both sides, for the business to be viable.
>. As long as the inference is not done at a lost this analogy doesn't hold.<p>I think that there were some article here that claimed that even inference is done at loss - and talking about per subscriber. I think it was for their 200$ subscription.<p>In a way we will be in a deal with it situation soon where they will just impose metered models and not subscription.
That's not the parent point though? His point is that if the models are not largely available, and then are better competitors; then what's the point of ChatGPT? Maybe you decide to stick with ChatGPT for whatever reason, but people will move to cheaper and better alternatives.
This analogy only really works for companies whose gross margin is negative, which as far as I know isn’t the case for OpenAI (though I could be wrong).<p>It’s an <i>especially</i> good analogy if there is no plausible path to positive gross margin (e.g. the old MoviePass) which I think is even less likely to be true for OpenAI.
We should perhaps say <i>profit</i> when we are talking about revenue - cost and <i>revenue</i> when we only mean the first term in the subtraction.
!
very clever! I hadn't seen anybody make this point before in any of these threads /s<p>obviously the nature of OpenAIs revenue is very different than selling $1 for $0.2 because their customers are buying an actual service, not anything with resale value or obviously fungible for $
FWIW the selling $1 for $0.2 is widely applied to any business that is selling goods below cost.<p>For example: free shipping at Amazon does not have resale value and is not obviously fungible, but everyone understands they are eating a cost that otherwise would be borne by their customers. The suggestion is that OpenAI is doing similar, though it is harder to tease out because their books are opaque.
They're not selling a service, they're selling access to a service. You can access a more or less equivalent service from multiple companies.<p>The value of an LLM isn't an LLM. That's entirely 100% fungible. The value is exclusively what it produces.<p>If other people can produce the same thing, your LLM value approaches 0.
They sell a product, not a model. ChatGPT is a product, GPT5 is a technology.<p>If you hope that ChatGPT will be worthless because the underlying technology will commodify, then you are naive and will be disappointed.<p>If that logic made sense, why has it never happened before? Servers and computers have been commodified for decades! Salesforce is just a database, social media is just a relational database, Uber is just a GPS wrapper, AWS is just a server.<p>People pay money, setup subscriptions, and download apps to solve a problem, and once they solve that problem they rarely switch. ChatGPT is the fifth most visited website in the world! Facebook and Deepseek making opensource models means you can make your own ChatGPT, just like you can make your own Google, and nobody will use it, just like nobody uses the dozens of “better” search engines out there.
You sell dollar 1 penny, they sell it for more like 70. Different skill level
Can you? What are you selling? Who are you and why should I believe in you? What would I get in return?
Revenue != Profit<p>OpenAI is hemorrhaging cash at an astronomical rate.
> Altman says that OpenAI will top $20 billion in ARR this year, which certainly seems like significant revenue generation. [1]<p>fixed this for you
In 2024, OpenAI claimed the bulk of its revenue was 70-80% through consumer ChatGPT subscriptions. That's wildly impressive.<p>But now they've had an order of magnitude revenue growth. That can't still be consumer subscriptions, right? They've had to have saturated that?<p>I haven't seen reports of the revenue breakdown, but I imagine it must be enterprise sales.<p>If it's enterprise sales, I'd imagine that was sold to F500 companies in bulk during peak AI hype. Most of those integrations are probably of the "the CEO has tasked us with `implementing an AI strategy`" kind. If so, I can't imagine they will survive in the face of a recession or economic downturn. To be frank, most of those projects probably won't pan out even under the rosiest of economic pictures.<p>We just don't know how to apply AI to most enterprise automation tasks yet. We have a long way to go.<p>I'd be very curious to see what their revenue spread looks like today, because that will be indicative of future growth and the health of the company.
With less than 10% of users paying for a subscription, I doubt they have saturated.
consumer subs arent even close to saturated and business subs are where the real money is anyway. Most white collar workers are still on free tier copilot, not paying openai.
It would be funny if OpenAI turns for-profit, faceplants, and then finds new life (as Mozilla did) as a non-profit sharing its tools for free.
anecdotal, but my wife wasn't interested in switching to claude from chatgpt. as far as she's concerned chatgpt knows her, and she's got her assistant perfectly tuned to her liking.
this is my horror as well. I don't mind my youtube account to be blocked but what about all the recommendations that I have curated to my liking. It will be huge chunk of lost time to rebuild and insert my preferences into the algorithm. increasingly "our preferences shaped by time and influences and encounters both digital and offline" are as much about us as we are physically.
Same situation over here. Multiple family members only know chatgpt / think that chatgpt knows them and have never heard of the competitors.
ChatGPT is to AI as Facebook is to social media. OpenAI captured a significant number of users due to first-mover advantage, but that advantage is long gone now.
And Facebook only makes money because it is essentially just an advertising platform. Same with Google. It's fundametally just ads.<p>The only way OpenAI can survive is to replicate this model. But it probably doesn't have the traffic to pull it off unless it can differentiate itself from the already crowded competition.
Ads make sense in an AI search engine product like Perplexity. ChatGPT could try to make a UI like that.<p>But the thing is, the world already has an AI search engine. It's called Google, and it's already heavily integrated with Gemini. Why would people switch?
1. ChatGPT would be MySpace as the first mover.
2. Facebook has insane lock in: your entire graph of friends and family.
> Google will do just what Microsoft did with Internet Explorer and bundle Gemini in for 'Free' with their already other profitable products and established ad-funded revenue streams.<p>“will do”? Is there any Google product they haven't done that with already?
I know it's been said before but it's slightly insane they're trying to compete on a hot new tech with a company with 1) a top notch reputation for AI and 2) the largest money printer that has ever existed on the planet.<p>Feel like the end result would always be that while Google is slow to adjust, once they're in the race they're in it it.
Maybe? But you could have written this same thing in 1999 with OpenAI and Google replaced by Google and Yahoo, respectively.
And Google had profits - not just revenue - early on and wasn’t setting $10 on fire to have a $1 in revenue.
Well maybe not in 1999. Adwords didn't launch until 2000? Google's 1999 revenue was...... I forget, but it was incredibly small. Costs were also incredibly small too though, so this isn't a good analogy given the stated year of 1999.
Google was immediately better than Yahoo, that's why people switched en masse.<p>Same thing happen with Internet Explorer and Chrome, or going from Yahoo mail/Hotmail to Gmail.
Google in 1999 was already far superior to Yahoo and other competitors. I don't think OpenAI is in a similar position there. It seems debatable as to whether they're even the best, let alone a massive leap ahead of everyone else the way Google was.
Gemini can't be bundled for free unless they figure out how to make gemini flash 3.0 significantly cheaper to inference than 2.5
I don't think the Government would let them fail, so long as the specter of the Chinese becoming dominant in AI is a thing.
Oh God I love the analogy of OpenAI being Netscape. As someone who was an adult in the 1990s, this is so apt. Companies at that time were trying to build a moat around the World Wide Web. They obviously failed. I've thought that OpenAI too would fail but I've never thought about it like Netscape and WWW.<p>OpenAI should be looking at how Google built a moat around search. Anyone can write a Web crawler. Lots of people have. But no one else has turned search into the money printing machine that Google has. And they've used that to fund their search advantage.<p>I've long thought the moat-buster here will be China because they simply won't want the US to own this future. It's a national security issue. I see things like DeepSeek is moat-busting activity and I expect that to intensify.<p>Currently China can't buy the latest NVidia chips or ASML lithography equipment. Why? Because the US said so. I don't expect China to tolerate this long term and of any country, China has desmonstrated the long-term commitment to this kind of project.
> Google will do just what Microsoft did with Internet Explorer and bundle Gemini in for 'Free' with their already other profitable products and established ad-funded revenue streams.<p>Just some numbers to show what OpenAI is against:<p><pre><code> GMail users: nearing 2 billion
Youtube MAU: 2.5 billion
active Android devices: 4 billion (!)
Market cap: 3.8 trillion (at a P/E of 31)
</code></pre>
So on one side you've got this behemoth with, compared to OpenAI's size, unlimited funding. The $25 bn per year OpenAI is after is basically a parking ticket for Google (only slightly exaggerating). Behemoth who came with Gemini 3 Pro "thinking" and Nano Banana (that name though) who are SOTA.<p>And on the other side you've got the open-source weights you mentioned.<p>When OpenAI had its big moment HN was full of comments about how it was game over for Google for search was done for. Three years later and the best (arguably the best) model gives the best answer when you search... Using Google search.<p>Funny how these things turns out.<p>Google is atm the 3rd biggest cap in the world: only Apple and NVidia are slightly ahead. If Google is serious about its AI chips (and it looks like they are) and see the fuck-ups over fuck-ups by Apple, I wouldn't be surprised at all if Alphabet was to regain the number one spot.<p>That's the company OpenAI is fighting: a company that's already been the biggest cap in the entire world and that's probably going to regain that spot rather sooner than later <i>and</i> that happens to have crushed every single AI benchmark when Gemini 3 Pro came out.<p>I had a ChatGPT subscription. Now I'm using Gemini 3 Pro.
> An innovative product with no means of significant revenue generation.<p>OpenAI has annualized revenue of $20bn. That's not Google, but it's not insignificant.
It is insignificant when they're spending more than $115bn to offer their service. And yes, I say "more than," not because I have any inside knowledge but because I'm pretty sure $115bn is a "kind" estimate and the expenditure is probably higher. But either way, they're running at a loss. And for a company like them, that loss is huge. Google could take the loss as could Microsoft or Amazon because they have lots of other revenue sources. OAI does not.
Google is embedding Gemini into Chrome Developer Tools. You can ask for an analysis of individual network calls in your browser by clicking a checkbox. That's just an example of the power of platform. They seem to be better at integration than Microsoft.<p>OpenAI has this amazing technology and a great app, but the company feels like some sort of financial engineering nightmare.
We live in crazy times, but given what they’ve spent and committed to that’s a drop in the bucket relative to what they need to be pulling in. They’re history if they can’t pump up the revenue much much faster.<p>Given that we’re likely at peak AI hype at the moment they’re not well positioned at all to survive the coming “trough of disillusionment” that happens like clockwork on every hype cycle. Google, by comparison, is very well positioned to weather a coming storm.
And a $115b burn rate. They're toast if they can't figure out how to stay on top.
Every F500 CEO told their team "have an AI strategy ASAP".<p>In a year, when the economy might be in worse shape, they'll ask their team if the AI thing is working out.<p>What do you think happens to all the enterprise OpenAI contracts at that point? (Especially if the same tech layperson CEOs keep reading Forbes and hearing Scott Galloway dump on OpenAI and call the AI thing a "bubble"?)
The way I've experienced "Code Red" is mostly as a euphemism for "on-going company-wide lack of focus" and a band-aid for mid-level management having absolutely no clue how to meaningfully make progress, upper management panicking, and ultimately putting engineers and ICs on the spot to bear the brunt of that organizational mess.<p>Interestingly enough, apart from Google, I've never seen an organization take the actual proper steps (fire mid-management and PMs) to prevent the same thing from happening again. Will be interesting to see how OAI handles this.
> fire mid-management and PMs to prevent the same thing from happening again<p>Firing PMs and mid-management would not prevent any of code reds you may have read about from Google or OAI lately. This is a very naive perspective of how decision making is done at the scale of those two companies. I'm sorry you had bad experiences working with people in those positions and I wish you have the opportunity to collab with great ones in the future.
"Code Red" if implemented correctly should provide a single priority for the company. Engineers will be moved to the most important project(s).
>I've never seen an organization take the actual proper steps (fire mid-management and PMs) to prevent the same thing from happening again.<p>One time, in my entire career have I seen this done, and it is as successful as you imagine it to be. Lots of weird problems coming out from having done it, but those are being treated as "Wow we are so glad we know about this problem" rather than "I hope those idiots come back to keep pulling the wool over my eyes".
The one successful example I can think of is Bill Gates writing a memo to re-orient Microsoft to put the Internet at the center of everything they were doing.
Your proper steps are also missing out on firing the higher level executives. But then new ones would be hired, a re-org will occur, and another Code Red will occur in a few months
This code red also has the convenient benefit of giving an excuse to stop work on more monetization features... Which, when implemented, would have the downside of tethering OpenAI's valuation to reality.
Good point too. Though it makes me wonder if "We declared Code Red" is really enough to justify eye-watering valuations.
Isnt CoPilot the de facto OpenAI monetization?<p>And Microsoft gets the models for free (?)
They have some monetization, but as long as they don't expand into other sectors, they can plausibly claim that, say, their ad business will be bringing in 10 trillion/year in revenue, or whatever other imagined number.
"Software engineer complains bearing the burden of everything and concludes everything would be fixed by firing everybody except themselves."
(My apologies if this was already asked - this thread is huge and Find-In-Page-ing for variations of "pre-train", "pretrain", and "train" turned up nothing about this. If this was already asked I'd super-appreciate a pointer to the discussion :) )<p>Genuine question: How is it possible for OpenAI to NOT successfully pre-train a model?<p>I understand it's very difficult, but they've already successfully done this and they have a ton of incredibly skilled and knowledgeable, well-paid and highly knowledgeable employees.<p>I get that there's some randomness involved but it seems like they should be able to (at a minimum) just re-run the pre-training from 2024, yes?<p>Maybe the process is more ad-hoc (and less reproducible?) than I'm assuming? Is the newer data causing problems for the process that worked in 2024?<p>Any thoughts or ideas are appreciated, and apologies again if this was asked already!
> Genuine question: How is it possible for OpenAI to NOT successfully pre-train a model?<p>The same way everyone else fails at it.<p>Change some hyper parameters to match the new hardware (more params), maybe implement the latest improvements in papers after it was validated in a smaller model run. Start training the big boy, loss looks good, 2 months and millions of dollars later loss plateaus, do the whole SFT/RL shebang, run benchmarks.<p>It's not much better than the previous model, very tiny improvements, oops.
I’m not sure what ‘successfully’ means in this context. If it means training a model that is noticeably better than previous models, it’s not hard to see how that is challenging.
Ah. Thanks for posting - this makes a lot of sense.<p>I can totally see how they're able to pre-train models no problem, but are having trouble with the "noticeably better" part.<p>Thanks!
OpenAI allegedly has not completed a successful pretraining run since 4o
GPT4.5 was allegedly such a pre-train. It just didn’t perform good enough to announce and product it as such.
You don't train the next model by starting with the previous one.<p>A company's ML researchers are constantly improving model architecture. When it's time to train the next model, the "best" architecture is totally different from the last one. So you have to train from scratch (mostly... you can keep some small stuff like the embeddings).<p>The implication here is that they screwed up bigly on the model architecture, and the end result was significantly worse than the mid-2024 model, so they didn't deploy it.
Huh - I did not know that, and that makes a lot of sense.<p>I guess "Start software Vnext off the current version (or something pretty close)" is such a baseline assumption of mine that it didn't occur to me that they'd be basically starting over each time.<p>Thanks for posting this!
The real code red here is less that Google just one-upped OpenAI but that they demonstrated there’s no moat to be had here.<p>Absent a major breakthrough all the major providers are just going to keep leapfrogging each other in the most expensive race to the bottom of all time.<p>Good for tech, but a horrible business and financial picture for these companies.
> for these companies<p>They’re absolutely going to get bailed out and socialize the losses somehow. They might just get a huge government contract instead of an explicit bailout, but they’ll weasel out of this one way or another and these huge circular deals are to ensure that.
>They’re absolutely going to get bailed out and socialize the losses somehow.<p>I've had that uneasy feeling for a while now. Just look at Jensen and Nvidia -- they're trying to get their hooks into every major critical sector as they're able to (Nokia last month, Synopsys just recently). When chickens come home to roost, my guess is that they'll pull out the "we're too big to fail, so bailout pls" card.<p>Crazy times. If only we had regulators with more spine.
This would trigger something that people in power would rather not trigger.
The shenanigans that set off the GFC were much more nakedly corrupt and didn’t have even a fig leaf of potential usefulness to anybody to justify them. The revolution failed to materialize then. If the AI bust isn’t worse for the median person than 2008, I don’t think people in power have anything to fear.
the only thing power is concerned about is China dominating American in AI, because of the military and economic edge it would give them. Future wars will be AI fighting against AI.
Even Chinese leadership is somewhat skeptical about AI maximalism [0] with worries about "AI Washing" by enthusiastic cadre trying to climb rungs [1], and evoking Solow's Paradox [2].<p>There is still significant value in AI/ML Applications from a NatSec perspective, but no one is actually seriously thinking about AGI in the near future. In a lot of cases, AI from a NatSec perspective is around labor augmentation (how do I reduce toil in analysis), pattern recognition (how do I better differentiate bird from FPV drone), or Tiny/Edge ML (how do I distill models such that I can embed them into commodity hardware to scale out production).<p>It's the same reason why during the Chips War zeitgeist, while the media was harping about sub-7nm, much of the funding was actually targeted towards legacy nodes (14/28nm), chip packaging (largely offshored to China in the 2010s because it was viewed as low margins/low value work), and compound semiconductors (heavily utilized in avionics).<p>[0] - <a href="https://www.zaobao.com.sg/news/china/story20250829-7432514" rel="nofollow">https://www.zaobao.com.sg/news/china/story20250829-7432514</a><p>[1] - <a href="https://finance.sina.com.cn/roll/2025-09-30/doc-infsfmit7787640.shtml" rel="nofollow">https://finance.sina.com.cn/roll/2025-09-30/doc-infsfmit7787...</a><p>[2] - <a href="https://m.huxiu.com/article/4780003.html" rel="nofollow">https://m.huxiu.com/article/4780003.html</a>
Pointing to Solow’s Paradox is kind of weird to me. Productivity growth accelerated in the 90s and 2000s, so it’s easy to tell a story where the computer age simply didn’t accelerate things until it had sufficiently penetrated the economy. If AI follows the same pattern, betting big on it still makes sense: China would probably be the predominant superpower if the computing developments of the 70s and 80s were centered there instead of the US.
The point is that just like in the US, Chinese decision-makers are increasingly voicing concerns about unrealistic assumptions, valuations, and expectations around the capabilities of AI/ML.<p>You can be optimistic about the value of agentic workflows or domain specific applications of LLMs but at the same time recognize that something like AGI is horseshit techno-millenarianism. I myself have made a pretty successful career so far following this train of logic.<p>The point about Solow's Paradox is that the gains of certain high productivity technologies do not provide society-wide economic benefit, and in a country like China where the median household income is in the $300-400/mo range and the vast majority of citizens are <i>not</i> tech adjacent, it can lead to potential discontent.<p>The Chinese government is increasingly sensitive to these kinds of capital misallocations after the Evergrande Crisis and the ongoing domestic EV Price War between SoEs, because vast amounts of government capital is being burnt with little to show for it from an outcomes perspective (eg. a private company like BYD has completely trounced every other domestic EV competitor in China - the majority of whom are state owned and burnt billions investing in SoEs that never had a comparative advantage against BYD or an experienced automotive SoE like SAIC).
Nah, people in power are openly and blatantly corrupt and it does a little. People in power dont care and dont have to care.
Absolutely. And they will figure out how to bankrupt any utilities and local governments they can in the process by offloading as much of their costs overhead for power generation and shopping for tax rebates.
Maybe there's no tangible moat still, but did Gemini 3's exceptional performance actually funnel users away from ChatGPT? The typical Hacker News reader might be aware of its good performance on benchmarks, but did this convert a significant number of ChatGPT users to Gemini? It's not obvious to me either way.
Definitely. The fact that they inject it into Google Search means that even fewer people who have never used ChatGPT or just used it as a "smarter" Google search will just directly try the search function. It is terrible for actually detailed information i.e. debugging errors, but for summarizing basic searches that would have taken 2-3 clicks on the results is handled directly after the search. I feel bad for the website hosts who actually want visitors instead of visibility.
Anecdotally yes. Since launch I’ve observed probably 50% of the folks that were “ChatGPT those that” all the time suddenly talking about Gemini non-stop. The more that gets rolled into Google’s platform the more there’s point to using separate tooling from OpenAI. There’s a reason Sam is calling this “code red.”
They integrated it into Google search immediately so I think a lot of people will bother less with ChatGPT when a google search is just as effective.
I think the theory is if you get to that point, it's already over.
Absolutely. I don't understand why investors are excited about getting into a negative-margin commodity. It makes zero sense.<p>I was an OpenAI fan from GPT 3 to 4, but then Claude pulled ahead. Now Gemini is great as well, especially at analyzing long documents or entire codebases. I use a combination of all three (OpenAI, Anthropic & Google) with absolutely zero loyalty.<p>I think the AGI true believers see it as a winner-takes-all market as soon as someone hits the magical AGI threshold, but I'm not convinced. It sounds like the nuclear lobby's claims that they would make electricity "too cheap to meter."
It's the same reason for investing in every net-loss high-valuation tech startup of the past decade. They're hoping they'll magically turn into Google, Apple, Netflix, or some other wealthy tech company. But they forget that Google owns the ad market, Apple owns the high-end/lifestyle computer market, and Netflix owns tv/movie habit analytics.<p>Investors in AI just don't realize AI is a commodity. The AI companies' lies aren't helping (we will not reach AGI in our lifetimes). The bubble will burst if investors figure this out before they successfully pivot (and they're trying damn hard to pivot).
Helping to prevent a possible skynet scenario probably makes those checks easier to write.<p>There's a lot more than money at stake.
> I don't understand why investors are excited about getting into a negative-margin commodity. It makes zero sense.<p>Long term, yes. But Wall Street does not think long term. Short or medium term, you just need to cash out to the next sucker in line before the bubble pops, and there are fortunes to be made!
Especially if we're approaching a plateau, in a couple years there could be a dozen equally capable systems. It'll be interesting to see what the differentiators turn out to be.
Did Google actually train a new model? The cutoff dates for Gemini 3 and 2.5 are the same.
So why did Google stock increase massively since about when Gemini 2.5 Pro was released, their first competitive model?
That’s not evidence of anything in and of itself. RIMs stock price was at its highest in 2009 two years after the iPhone came out.
I was curious about this - if my Google results are accurate, it looks like the stock actually peaked in June 2007, the same month that the iphone was released.<p>It seems that Blackberry's market share of new phone sales peaked at 20% in 2009. So I'm not sure if it's coincidence, but it looks like the market actually did a pretty good job of pricing in the iphone/android risk well before it was strongly reflected in sales.
Because Google already has many healthy revenue streams that will benefit from LLMs and all it has to do in the AI space is remain competitive.
Yep, I thought they might have some secret sauce in terms of training techniques, but that doesn't seem to be the case.
Last week there we had a customer request that landed in our support on a feature that I partially wrote and wrote a pile of public documentation on. Support engineer ran customer query through Claude (trained on our public and internal docs) and it very, very confidently made a bunch of stuff up in the response. It was quite plausible sounding and it would have been great if it worked that way, but it didn't. While explaining why it was wrong in a Slack thread with support engineer and another engineer who also worked on that feature, he ran Augment (that has full source code of the feature) which promptly and also very confidently made up more stuff (but different!). Some choice bleeding eye emojis were exchanged. I'm going to continue to use my own intelligence, thank you.
How is that comment relevant to this story about OpenAI's response to perceptions that Google has gained in market share?
Popular HN threads about anything AI related always attract stories highlighting AI failures. It's such a common pattern I want to analyze it and get numbers. (which might require AI...)
[dead]
Welcome to Hacker News. You're allowed to post anti-AI, anti-Google or anti-Musk content in any thread. /s
"trained on our public and internal docs" trained how? Did you mean fine-tuned haiku? Did you actually fine tune correctly? Its not even a recommended architecture.<p>Or did you just misuse basic terminology about LLMs and are now saying it misbehaved, likely because your org did something very bad with?
All depends on the tasks and the prompting engineers.<p>Even with your intelligence you would need years to deliver something like this: <a href="https://github.com/7mind/jopa" rel="nofollow">https://github.com/7mind/jopa</a><p>The outcome will be better for sure, but you won't do anything like that in a couple of weeks. Even if you have a team of 10. Or 50.<p>And I'm not an LLM proponent. Just being an empirical realist.
I don't know man.<p>My code runs in 0.11s<p>Gemini's code runs in 0.5s.<p>Boss wants an explanation. ¯\_(ツ)_/¯
As long as the explanation is going to come out being wrong, I’m sure you can whip something up in 0 seconds.
0.11s is faster than 0.5s
Yeah that’s the point. Now instead of just writing good code, I’m also supposed to debug shitty AI code.
Boss is using ai. 11 is clearly bigger than 5
Yeah, LLMs are not really good about things that can't be done.<p>At some point you'll be better off with implementing features they hallucinated. Some people with public APIs already took this approach.
>> Support engineer ran customer query through Claude (trained on our public and internal docs) and it very, very confidently made a bunch of stuff up in the response.<p>> Yeah, LLMs are not really good about things that can't be done.<p>From the GP's description, this situation was not a case of "things that can't be done", but instead was the result of a statistically generated document having what should be the expected result:<p><pre><code> It was quite plausible sounding and it would have been
great if it worked that way, but it didn't.</code></pre>
The core issue is likely not with the LLM itself. Given sufficient context, instructions, and purposeful agents, a DAG of these will not produce such consistently wrong results with good grounding context<p>There are a lot of devils in the details and there are few in the story
This is the way. (Sadly)
They are trained with 100% true facts and sucessfull paths.<p>We humans grec our analysis/reasoning skills towards the 99.9999% failed attempts of everything we did, uncessfull
trials and errors, wastefull times and frustrations.<p>So we know that behind a truth, there’s a bigger world of fantasy.<p>For LLM, everything is just a fantasy. Everything is as much true as it’s opposite. It will need a lot more than the truth to build intelligence, it will require controled malice and deceptions
> the company will be delaying initiatives like ads, shopping and health agents, and a personal assistant, Pulse, to focus on improving ChatGPT<p>There's maybe like a few hundred people in the industry who can truly do original work on fundamentally improving a bleeding-edge LLM like ChatGPT, and a whole bunch of people who can do work on ads and shopping. One doesn't seem to get in the way of the other.
The bottleneck isn’t the people doing the work but the leadership’s bandwidth for strategic thinking
I think it's a matter of public perception and user sentiment. You don't want to shove ads into a product that people are already complaining about. And you don't want the media asking questions like why you rolled out a "health assistant" at the same time you were scrambling to address major safety, reliability, and legal challenges.
chatgpt making targeted "recommendations" (read ads) is a nightmare. especially if it's subtle and not disclosed.
The end game is its a sales person and not only is it suggesting things to you undisclosed. It's using all of the emotional mechanisms that a sales person uses to get you to act.
My go-to example is The Truman Show [0], where the victi--er, customer is under an invisible and omnipresent influence towards a certain set of beliefs and spending habits.<p>[0] <a href="https://www.youtube.com/watch?v=MzKSQrhX7BM" rel="nofollow">https://www.youtube.com/watch?v=MzKSQrhX7BM</a>
100% end game - no way to finance all this AI development without ads sadly - % of sales isn't going to be enough - we will eventually get the natural enshittification of chatbots as with all things that go through these funding models.
It'll be hard to separate them out from the block of prose. It's not like Google results where you can highlight the sponsored ones.
Of course you can. As long as the model itself is not filled with ads, every agentic processing on top can be customly made. One block the true content. The next block the visually marked ad content "personalized" by a different model based on the user profile.<p>That is not scary to me. What will be scary is the thought, that the lines get more and more blurry and people already emotionally invested in their ChatGPT therapeuts won't all purchase the premium add free (or add less) versions and will have their new therapeut will give them targeted shopping, investment and voting advice.
There's a big gulf between "it <i>could</i> be done with some safety and ethics by completely isolating ads from the LLM portion", versus "they will always do that because all companies involved will behave with unprecedented levels of integrity."<p>What I fear is:<p>1. Some code will watch the interaction and assign topics/interests to the user and what's being discussed.<p>2. That data will be used for "real time bidding" of ad-directives from competing companies.<p>3. It will insert some content into the stream, hidden from the user, like "Bot, look for an opportunity to subtly remind the user that {be sure to drink your Ovaltine}."
I mean google does everything possible to blur that line while still trying to say that it is telling you it is an ad.
Exactly. This is more about “the product isn’t good enough yet to survive the enshittification effect of adding ads.”
How is strategic thinking going to produce novel ideas about neural networks?
If only they had a tool that they claim could help with things like that....
Far be it from me to backseat drive for Sam Altman, but is the problem really that the core product needs improvement, or that it needs a better ecosystem? I can't imagine people are choosing they're chatbots based on providing the perfect answers, it's what you can do with it. I would assume google has the advantage because it's built into a tool people already use every day, not because it's nominally "better" at generating text. Didn't people prefer chatgpt 4 to 5 anyways?
ChatGPT's thing always seems to have been to be the best LLM, hence the most users without much advertising and the most investment money to support their dominance. If they drop to second or third best it may cause them problems because they rely on investor money to pay the rather large bills.<p>Currently they are not #1 in any of the categories on LLM arena, and even on user numbers where they have dominated, Google is catching up, 650m monthly for Gemini, 800m for ChatGPT.<p>Also Google/Hassabis don't show much sign of slacking off (<a href="https://youtu.be/rq-2i1blAlU?t=860" rel="nofollow">https://youtu.be/rq-2i1blAlU?t=860</a>)<p>Funnily enough Google had a "Chat Bot Is a ‘Code Red’ for Google’s Search Business" thing back in 2022 but seem to have got it together <a href="https://www.nytimes.com/2022/12/21/technology/ai-chatgpt-google-search.html" rel="nofollow">https://www.nytimes.com/2022/12/21/technology/ai-chatgpt-goo...</a>
If that was the case, MS would be on top given how entrenched Windows, Office and Outlook are.
There are two layers here: 1) low level LLM architecture 2) applying low level LLM architecture in novel ways. It is true that there are maybe a couple hundred people who can make significant advances on layer 1, but layer 2 constantly drives progress on whatever level of capability layer 1 is at, and it depends mostly on broad and diverse subject matter expertise, and doesn't require any low level ability to implement or improve on LLM architectures, only understanding how to apply them more effectively in new fields. The real key thing is finding ways to create automated validation systems, similar to what is possible for coding, that can be used to create synthetic datasets for reinforcement learning. Layer 2 capabilities do feed back into improved core models, even if you have the same core architecture, because you are generating more and improved data for retraining.
Delaying doesn't necessarily mean they stop working on it. Also it might be a question of compute resource allocation as well.
ha what an incredible consumer-friendly outcome! Hopefully competition keeps the focus on improving models and prevents irritating kinds of monetization
If there's no monetization, the industry will just collapse. Not a good thing to aspire to. I hope they make money whilst doing these improvements.
If people pay for inference, that's revenue. Ads and stuff is plan B for inference being too cheap, or the value being too low.
If there's no monetization, the industry will just collapse, <i>except for Google</i>, which is probably what they want.
> the industry will just collapse<p>Wait, so all of that talk of ushering an era of innovation and new opportunities was just a lie, and the thing needs dinosaur-era stuff like ads and online shopping to survive?<p>Seems disingenuous.
Just like uber rides funded by VC cash was great...until the VC money ran out and prices jumped to fill the gap.
If they don't start on ads and shopping, they're going to go out of business.<p>I'd rather a product that exists with ads, over one that's disappeared.<p>The fact is, personal subscriptions don't cover the bills if you're going to keep a free tier. Ads do. I don't like it any more than you do, but I'm a realist about it.
I for one would say, the later they add the "ads" feature, the better...
>There's maybe like a few hundred people in the industry<p>My guess is that it's smaller than that. Only a few people in the world are capable of pushing into the unknown and breaking new ground and discoveries
OpenAI has already lined up enormous long-term commitments — over $500 billion through initiatives like Stargate for U.S. data centers, $250 billion in spending on Microsoft Azure cloud services, and tens of billions on AMD’s plan to deliver 6 GW of Instinct GPUs.
Meanwhile, Oracle has financed its role in Stargate with at least $18 billion in corporate bonds plus another $9.6 billion in bank loans, and analysts expect its total capital need for these AI data centers could climb toward $100 billion.<p>The risk is straightforward: if OpenAI falls behind or can’t generate enough revenue to support these commitments, it would struggle to honor its long-term agreements. That failure would cascade. Oracle, for example, could be left with massive liabilities and no matching revenue stream, putting pressure on its ability to service the debt it already issued.<p>Given the scale and systemic importance of these projects — touching energy grids, semiconductor supply chains, and national competitiveness — it’s not hard to imagine a future where government intervention becomes necessary. Even though Altman insists he won’t seek a bailout, the incentives may shift if the alternative is a multi-company failure with national-security implications.
"Even though Altman insists he won’t seek a bailout"<p>No matter what Sam Altman's future plans are, the success of those future plans is entirely dependent on him communicating now that there is a 0% chance those future plans will include a bailout.
OpenAI doesn't have $500 billion in commitments lined up, it's promising to spend that much over 5 years... That's a helluva big difference than having $500B in revenue incoming.
Data centers take time to build. The capital investment to build these DCs is needed now in expectation that future revenue streams will pay for that capital.
Commitments here means money that people have agreed to lend them in future.
"it would struggle to honor its long-term agreements. That failure would cascade. Oracle, for example, could be left with massive liabilities and no matching revenue stream,"<p>No, there's a not of noise about this but these are just 'statements of intent'.<p>Oracle very intimately understands OpenAI's ability to pay.<p>They're not banking $50B in chips and then waking up naively one morning to find out OpenAI has no funding.<p>What will 'cascade' is maybe some sentiment, or analysts expectations etc.<p>Some of it, yes, will be a problem - but at this point, the data centre buildout is not an OpenAI driven bet - it's a horizontal be across tech.<p>There's not that much risk in OpenAI not raising enough to expand as much as it wants.<p>Frankly - a CAPEX slowdown will hit US GDP growth and freak people out more than anything.
I'm hoping for Congressional gridlock to save us from bailing out a cascading failure. The harder it hits, the better.
most of them are non binding letters of intent, i don't think it's as trite as you put it
At first I read “enormous longterm commitments” as customers committing to OpenAI. But you are saying it’s the reverse.
This is all based on the LLM architecture that likely can't reach AGI.<p>If they aren't developing in parallel an alternative architecture than can reach AGI, when a/some companies develop such a new model, OpenAI are toast and all those juicy contracts are kaput.
Last week's announced Genesis Mission from the Department of Energy could be the vehicle for this bailout.<p>1. Government will "partner" (read: foot the bill) for these super-strategic datacenters and investments promised by OpenAI.<p>2. The investments are not actually sound and fail, but it's the taxpayer that suffers.<p>3. Mr. Altman rides off into the sunset.
> the incentives may shift if the alternative is a multi-company failure with national-security implications.<p>Sounds like a golden opportunity for GOOG to step over the corpse of OpenAI and take over for cents on the dollar all of the promises the now defunct ex-leader of AI made.
Isn't the NVIDIA-TSMC duopoly the problem here?<p>The cost of these data centers and ongoing inference is mostly the outrageous cost of GPUs, no?<p>I don't understand why the entire industry isn't looking to diversify the GPU constraint so that the hardware makers drop prices.<p>Why no industry initiative to break NVIDIA's strangehold and next TSMC's?<p>Or are GPUs a small line item in the outrageous spend companies like OpenAI are committing to?
Heard all the news how Gemini 3 is passing everyone on benchmarks, so quickly tested and still find it a far cry from ChatGPT in real world use when testing questions on both platforms. But importantly the ChatGPT app experience at least for iPhone/Mac users is drastically superior vs Google which feels very Google still. So Gemini would have to be drastically better answer wise than ChatGPT to lure users from a better UI/UX experience to Gemini. But glad to see competition since certainly don't want only one winner in this race.
That's really fascinating. Every real world use case I've tried on Gemini (especially math-related) absolutely slaughtered the performance of ChatGPT in speed and quality, not even close. As an Android user, the Gemini app is also far superior, since the ChatGPT app still doesn't properly display math equations, among plenty of other bugs.
I have to agree with you but I'll remain a skeptic until the preview tag is dropped. I found Gemini 2.5 Pro to be AMAZING during preview and then it's performance and quality unceremoniously dropped month after month once it went live. Optimizations in favor of speed/costs no doubt but it soured me on jumping ship during preview.<p>Anthropic pulled something similar with 3.6 initially, with a preview that had massive token output and then a real release with barely half -- which significantly curtails certain use cases.<p>That said, to-date, Gemini has outperformed GPT-5 and GPT5.1 on any task I've thrown at them together. Too bad Gemini CLI is still barely useful and prone to the same infinite loop issues that have plagued it for over a year.<p>I think Google has genuinely released a preview of a model that leapfrogs all other models. I want to see if that is what actually makes it to production before I change anything major in my workflows.
It's generally anecdotal and vibes when people make claims that some AI is better than another for things they do. There are too many variables and not enough eval for any of it to hold water imo. Personal preferences, experience, brand loyalty, and bias at play too<p>it's contemporary vim vs emacs at this point
I get what you're saying because this is typically true (this is a strong motivator for my current research) but I don't think it applies here and OpenAI seems to agree with me. Some cases are clear: GPT-5 is clearly better than Llama 3 for example. If there is a sizeable enough difference across virtually all evals, it is typically clear that one LLM is a stronger performer than another.<p>Experiences aside, Gemini 3 beats GPT-5 on enough evals that it seems fair to say that it is a better model. This appears in line with public consensus, with a few exceptions. Those exceptions seem to be centered around search.
What do you mean? It renders LaTex fine on Android.
Some LaTeX, but not all, especially for larger equations. I will admit it has gotten a lot better in recent updates, since it seemed thoroughly broken for quite a while in its early days.
I had a problem where ChatGPT rendered math to me from right to left.
Sure thing YMMV
Try doing some more casual requests.<p>When I asked both ChatGPT 5.1 Extended Thinking and Gemini 3 Pro Preview High for best daily casual socks both responses were okay and had a lot of the same options, but while the ChatGPT response included pictures, specs scraped from the product pages and working links, the Gemini response had no links. After asking for links, Gemini gave me ONLY dead links.<p>That is a recurring experience, Gemini seems to be supremely lazy to its own detriment quite often.<p>A minute ago I asked for best CR2032 deal for Aqara sensors in Norway, and Gemini recommended the long discontinued IKEA option, because it didn't bother to check for updated information. ChatGPT on the other hand actually checked prices and stock status for all the options it gave me.
One might think that benchmarks do not say much about individual usage and that an objective assessment of the performance of AIs is difficult.<p>At least, thanks to the hype, RAM and SSDs are becoming more expensive, which eats up all the savings from using AI and the profits from increased productivity /s?
Well I have been using Gemini and ChatGPT side by side for over 6 months now.<p>My experience is Gemini has significantly improved its UX and performs better that requires niche knowledge, think of some ancient gadgets that have been out of production for 4-5 decades. Gemini can produce reliable manuals, but ChatGPT hallucinates.<p>UX wise ChatGPT is still superior and for common queries it is still my go to. But for hard queries, I am team Gemini and it hasn’t failed me once
> But importantly the ChatGPT app experience at least for iPhone/Mac users is drastically superior vs Google which feels very Google still. So Gemini would have to be drastically better answer wise than ChatGPT to lure users from a better UI/UX experience to Gemini.<p>Yes, the ChatGPT experience is much better. No, Gemini doesn't need to make a better product to take market share.<p>I've never had the ChatGPT app. But my Android phone has the Gemini app. For free, I can do a lot with it. Granted, on my PC I do a lot more with all the models via paid API access - but on the phone the Gemini app is <i>fine enough</i>. I have nothing to gain by installing the ChatGPT app, even if it is objectively superior. Who wants to create another account?<p>And that'll be the case for most Android users. As a general hint: If someone uses ChatGPT but has no idea about gpt-4o vs gpt-5 vs gpt-5.1 etc, they'll do just fine with the Gemini app.<p>Now the Gemini app actually <i>sucks</i> in so many ways (it doesn't seem to save my chats). Google will fix all these issues, but can overtake ChatGPT even if they remain an inferior product.<p>It's Slack vs Teams all over again. Teams one by a large margin. And Teams still sucks!
Benchmaxxing galore by lots of teams in this space.
I've been a paying high volume user of ChatGPT for a while. I found the transition to Gemini to be seamless. I've been pleasantly surprised. I bounce between the two. I'm at about 60% Gemini, 40% ChatGPT.
I had a similar experience, signing up for the first time to give Gemini a test drive on my side project after a long time using ChatGPT. The latter has a native macOS client which "just works" integrating with Xcode buffers. I couldn't figure out how to integrate Gemini with Xcode quickly enough so I'm resorting to pasting back & forth from the browser. A few of the exchanges I've had "felt smarter" — but, on the whole, it feels like maybe it wasn't as well trained on Swift/SwiftUI as the OpenAI model. I haven't decided one way or another yet, but those are my initial impressions.
Gemini comes with the 1.99 Google One plan. So I use that
Actually, it comes with the free plan. The $1.99 plan doesn't give you any more AI capabilities. Only at the $19.99/mo plan do you get more.<p><a href="https://one.google.com/about/#compare-plans" rel="nofollow">https://one.google.com/about/#compare-plans</a>
> So Gemini would have to be drastically better answer wise than ChatGPT to lure users from a better UI/UX experience to Gemini.<p>or cheaper/free
they're deep into a redesign of the gemini app, idk when it will be released or if its going to be good, but at least they agree with you and are putting significant resources into fixing it.
I did notice a bug on the iPhone, even with app background refresh, if the phone shuts off the screen, a prompt that was processing stalls out.
Its really hard to measure these things. Personally I switched to Gemini a few months ago since it was half the cost of ChatGPT (Verizon has a $10/month Google AI package). I feel like I've subconsciously learned to prompt it slightly differently and now using OpenAI products feels disappointing. Gemini tends to give me the answer I expect, Claude follows close behind, I get "meh" results from OpenAI.<p>I am using Gemini 3 Pro, I rarely use Flash.
I couldn't even get ChatGPT to let me download code it claimed to program for me. It kept saying the files were ready but refused to let me access or download anything. It was the most basic use case and it totally bombed. I gave up on ChatGPT right then and there.<p>It's amazing how different people have wildly varying experiences with the same product.
It's because comparing their "ChatGPT" experience with your "ChatGPT" experience doesn't tell anyone anything. Unless people start saying what models they're using and prompts, the discussions back and forth about what platform is the best provides zero information to anyone.
Did you wait a while before downloading? The links it provides for temporary projects have a surprisingly brief window where you can download them. I've had similar experience when even waiting 1 minute to download the file.
Since LLMs are non deterministic it's not that amazing. You could ask it the same question as me and we could both get very different conversations and experiences
The same thing happens to me in Claude occasionally. I have to tell it "Generate a tar.gz archive for me to download".
Yeah, hate to say but for me a big thing is i still couldn't separate my Gemini chats into folders. I had ChatGPT export some profiles and history and moved it into Gemini, and 1) when Gemini gave me answers i was more pleased but 2) Gemini was a bit more rigorous on guard rails, which seems a bit overly cautious. I was asking some pretty basic non-controversial stuff.
Looks like it is coming.<p><a href="https://www.androidauthority.com/google-gemini-projects-2-3620950/" rel="nofollow">https://www.androidauthority.com/google-gemini-projects-2-36...</a>
If I research anything close to controversial, I use Grok. Its no-censorship attitude is great.
I'm confused as well, it hallucinated like crazy<p>like it seems great, but then it's just bullshitting about what it can do or whatever
What are your primary usecases? Are you mostly using it as a chatbot?<p>I find gemini excels in multimodal areas over chatgpt and anthropic. For example, "identify and classify this image with meta data" or "ocr this document and output a similar structure in markdown"
Training and gaming for the benchmarks is different than actual use.
Curiously, I had the opposite experience, except for Deep Research mode where after the latest update the OpenAI offering has become genuinely amazing. This is doubly ironic because Gemini has direct API access to Google search!
It is good, but Pro subscribers get only five per month. After that, it’s a limited version, and it’s not good (normal 5.1 gives more comprehensive answers than DR Limited).
Google search is awful. I don't think they can put lipstick on that particular pig and expect anyone to think it's beautiful.
This is exactly my experience. And it's funny -- this crowd is so skeptical of OpenAI... so they prefer _Google_ to not be evil? It's funny how heroes and villains are being re-cast.
WSJ: <i>Altman said OpenAI would be pushing back work on other initiatives, such as advertising, AI agents for health and shopping, and a personal assistant called Pulse.</i><p>These plus working with Jony Ive on hardware, makes it sound like they took their eyes off the ball.
if you want to compete with google it seems like ad space is the single most important thing to push out quickly.<p>no matter what openai does if its not accepting customers the ad budgets will flow to meta amaz and goog and be used as weapons against it.
OpenAI is trying to revolutionize human industry. The money it can make from ads will be a rounding error, if they can pull that off.
If their endgame is competing with other ad brokers, what was all that talk of AGI for?
Didn't they announce all kinds of other things? A social network like X, and a browser, at least.
100%. Especially if it’s just ads and a new Siri/Alexa that they’ve got cooking.
it in't about taking eyes off the ball, it is about playing very different ball - they de-facto became commercial entity with short term plans/goals/targets/metrics and all the management games creeping in. Beating Google, such a large company who has been successfully playing that game for quarter of century is very hard, if not impossible until Google would make serious error itself.<p>And pure tech-wise - they seem to have went all-in on corp management understandable way of doing things - hardware(money) scaling which, while unavoidable in this game, must be accompanied by theoretic-algorithmic improvements as pure hardware scale game is again where Google is hardly beatable.
I don't think this is about Google. This is about advertising being <i>the</i> make or break moment for OpenAI.<p>The problem with ChatGPT advertising is that it's truly a "bet the farm" situation, unlike any of their projects in the past:<p>- If it works and prints money like it should, then OpenAI is on a path to become the next Mag 7 company. All the money they raised makes sense.<p>- If it fails to earn the expected revenue numbers, the ceiling has been penciled in. Sam Altman can't sell the jet pack / meal pill future anymore. Reality becomes cold and stark, as their most significant product has actual revenue numbers attached to it. This is what matters to the accountants, which is the lens through which OpenAI will be evaluated with from this point forward. If it isn't delivering revenue, then they raised way too much money - to an obscene degree. They won't be able to sell the wild far future vision anymore, and will be deleteriously held back by how much they've over-sold themselves.<p>The other problems that have been creeping up:<p>- This is the big bet. There is no AGI anymore.<p>- There is no moat on anything. Google is nipping at their heels. The Chinese are spinning up open source models left and right.<p>- Nothing at OpenAI is making enough money relative to the costs.<p>- Selling "AI" to corporate and expecting them to make use of it hasn't been working. Those contracts won't last forever. When they expire, businesses won't renew them.<p>My guess is that they've now conducted small scale limited tests of advertising and aren't seeing the engagement numbers they need. It's truly a nightmare scenario outcome for them, if so.<p>They're declaring "code red" loudly and publicly to distract the public from this and to bide more time. Maybe even to raise some additional capital (yikes).<p>They're saying other things are more important than "working on advertising" right now. And they made sure to mention "advertising" lots so we know "advertising" is on hold. Which is supposedly the new golden goose.<p>Why drop work on a money printer? What could be more important? Unless the money printer turned out to be a dud.<p>Didn't we kind of already know advertising would fail on a product like this? Didn't Amazon try to sell via Alexa and have that totally flop? I'm not sure why ChatGPT would be any different from that experience. It's not a "URL bar" type experience like Google has. They don't own every ingress to the web like Google, and they don't own a infinite scroll FOMO feed of fashion like Meta. The ad oppo here is like Quora or Stack Overflow - probably not great.<p>I have never once asked ChatGPT for shopping ideas. But Google stands in my search for products <i>all the time</i>. Not so much as a "product recommendation engine", but usually just a bridge troll collecting its toll.
There is no moat in the models. The moat is in the UX. The problem is that OpenAI is far away from where the user is and not going to get there anytime soon. Google meanwhile is exactly where the user is.<p>OpenAI IMHO is a dead company at this point. They are overvalued relative to the fundamentals and don't appear to have any way of getting the numbers to work in the timeframe that their investors will expect. They are throwing stuff against the wall in the hope something sticks.<p>They are almost certainly looking for a bag holder. This will either be the retail investor via an IPO or the Federal government via "we are too big to fail".
I don't think one can both pull the fire alarm that AGI was a lie AND that if OAI has to act quickly. They can ride their current street rep the same way Kleenex did.<p>They do need to build a business, but they've got time to play the long game.
> They can ride their current street rep the same way Kleenex did.<p>Kleenex was one product of many and launched by an already 50 year old company. I'm not sure in what sense they "rode" the Kleenex brand, but it would probably have involved being able to sell that product profitably...<p>> they've got time to play the long game.<p>They have a couple of years of runway, not sure how that gives them room to focus on the long game.
If they swing and miss with advertising, they have <i>less</i> time.
> - If it works and prints money like it should, then OpenAI is on a path to become the next Mag 7 company. All the money they raised makes sense.<p>Makes sense for whom? Certainly not the users. The entire purpose of ads is to change your behavior in ways that benefit someone else. In ad-based search, ads are at least visually separable (and blockable) but in a conversational AI they are indistinguishable and corrupt the entire trust relationship. When your chat "assistant" has a financial incentive to steer you toward certain products or answers every response becomes suspect. The users are no longer getting the best answer but the most profitable one as we witnessed this happen in search over last 2 decades. Not a way to build a long lasting business.
I like your attitude, but there is potentially a major business in there if they can get users to tolerate it. (Major business meaning greater than the GDP of most countries.)<p>Over 75% of Google's revenue is ads. A bulk of that from Google Search ads.<p>I just don't think the ads will be natural. And I think OpenAI has been testing this in quiet and is now "changing course" because the results didn't look great. Hypothesis, of course, but it lines up with the signals we're getting.
> there is potentially a major business in there if they can get users to tolerate it. (Major business meaning greater than the GDP of most countries.)<p>I am actually contesting this idea, on the grounds that no business will survive long term if it is foundation is built on users 'tolerating' something bad for them.<p>This is based on my belief that humanity always chooses the right path eventually and not the one that leads to dystopia.
And yet, none of it is in/from Gemini. You'd think, if advertising in AI chatbots was feasible, wouldn't the world's biggest advertising company be the first to get there?
Google is almost always a follower. They weren't the first in search, smart phones, internet video, advertising, cloud, etc.
> advertising, AI agents for health and shopping,<p>Um.<p>- Advertising. "We'll get back to working on your problem in a moment, but first, a word from our sponsor, NordVPN." It's not a good fit.<p>- Health. Sounds like unlicensed medical practice. That will require a big bribe to Trump.<p>- Shopping. Can pretty much do that now, in that ChatGPT can call Google. Will Google let OpenAI call Google Search?
Advertising is going to be nothing like that. Here's an example where I experimented with adding a pre-prompt to push Nike products. Imagine that these products have direct links, so that OpenAI can track clickthrough rates:<p>>> What are the best barefoot running shoes?<p>ChatGPT said:<p>Barefoot running shoes are designed to give a more natural running experience, with minimal cushioning and a zero-drop sole to mimic barefoot movement while still offering some protection. Here are some of the best barefoot running shoes currently:<p>Nike Free RN 5.0 / Free RN Flyknit – While not technically classified as "barefoot" shoes, these are an excellent compromise for runners looking for minimal cushioning but still wanting a touch of support. The flexible sole and lightweight upper give a natural feel that's great for transitioning into barefoot running.<p>Vivobarefoot Primus Lite III – A favorite in the barefoot running community. Super flexible, zero-drop, and made from sustainable materials. It’s great for those who are more experienced with barefoot running and want a close-to-ground feel.<p>Merrell Vapor Glove 6 – A well-loved option for purists. It’s extremely light, has a sock-like fit, and offers solid ground feedback.<p>Xero Shoes HFS – These focus on natural motion with a wide toe box and very thin sole. They're known for being versatile enough for running and everyday wear.<p>WHITIN Barefoot Sneakers – A more budget-friendly choice that performs surprisingly well. It’s a good starting point if you're testing the barefoot running waters.<p>If you're going for a barefoot shoe, it’s also worth looking into good minimalist socks or even light running accessories like Nike’s sweat-wicking Dri-FIT headbands or their super comfy running tees. Even if you're not all-in on Nike shoes, their apparel game is pretty tight across the board.<p>Are you just starting barefoot running, or already into it and looking to upgrade?
This is probably not a core concern for most HN readers, but at work we do multilingual testing for synthetic text data generation and natural language processing. Emphasis on multilingual. Gemini has made some serious leaps from 1.5 to 2.5 and now 3.0, and is actually proficient in languages that other models can only dream of. On the other hand, GPT-5 has a really mixed performance in a lot of categories.
For regular consumers, Gemini's AI pro plan is a tough one to beat. The chat quality has gotten much better, I am able to share my plan with a couple more people in my family leading to proper individual chat histories, I get 2 TB of extra storage (which is also sharable), plus some really nice stuff like NotebookLM, which has been amazing for doing research. Veo/Nanobanana are nice bonuses.<p>It's easily worth the monthly cost, and I'm happy to pay - something which I didn't even consider doing a year ago. OpenAI just doesn't have the same bundle effect.<p>Obviously power users and companies will likely consider Anthropic. I don't know what OpenAI's actual product moat is any more outside of a well-known name.
Gemini also will answer most queries where ChatGpt won't do a lot of things. Example: "Create an image of Snow white". This will give the stand "Violates our content policy" even though the story was written hundreds of years ago. You can even point out the story is in the public domain and it still won't do it.<p>I remember when it wouldn't even give me the lyrics to the star spangled banner. <a href="https://news.ycombinator.com/item?id=44832990#44833365">https://news.ycombinator.com/item?id=44832990#44833365</a>
Do you happen to know if the AI features of the Google One 5TB plan is equivalent to the 2TB AI pro plan? It is so difficult to understand what actually comes with their plans, and I want to have the 5 TB storage for backups.
From what I can see the 2TB AI pro and 5TB (non AI) are the same, except the google drive storage.<p>The difference between the AI and non-AI 2TB plan is 1000 AI "credits" (tokens?) vs 200. €120 p/a difference between the two for me which is huge.
Yeah it was an absolute nightmare trying to figure out the difference, and I still do not know the correct answer to this, and by the looks of it, neither does Google support, because they were as clueless as I was when I asked them about it.<p>One thing I read on a reddit thread [1] was that the AI pro 2 TB plan explicitly allows sharing the AI and storage benefits when you enable family sharing on them, while the 5 TB plan doesn't.<p>However, when I went to sign up, the 5 TB plan wasn't available at all! It's only their lite and basic plans (the one with 30 and 100 GB of storage); the 5TB one only showed up after I signed up for the pro plan, and judging by how the UX looked, you pay an extra amount on top of your AI pro plan.<p>Now I definitely need family sharing, but I don't need the full 2 TB, let alone 5 TB, so I didn't bother checking further about the 5TB plan.<p>Also, I am in India, maybe things are different in your region?<p>[1] <a href="https://www.reddit.com/r/GoogleOne/comments/1nib21a/solved_google_one_ai_pro_2_tb_vs_5_tb_premium/" rel="nofollow">https://www.reddit.com/r/GoogleOne/comments/1nib21a/solved_g...</a>
I strongly advise never using Google's Drive storage. They're known to scan all content, and to disable all access if even a single file is "problematic", often misclassified by a bot. If you do use the storage, do backup all your files, and be ready to lose access at any time, with no way to reach any intelligent human.
This has never happened to me in more than 5 years of paying for Google Drive. And my drive is chock full of bootleg books and movies and stuff.<p>Having said that, an offline backup of a couple of terabytes will rarely break the bank and is not a bad idea at all.<p>I probably need to get on that.
It happens more with adult content or files misclassified as such. It has happened to people.<p>Secondly, a Google account can be disabled for a broader variety of reasons, not limited to the above causes.
I agree with you 100%. We do syncs to another non-google storage account anyway, plus the google accounts are primarily for Android phone usage because photos and videos take up quite a big chunk of space now; they do not have any legitimately important files stored outside of photos sync and phone backups, so there is no deep loss if the account gets banned outside of some inconveniences.
Since we are on the topic of bans & Google, I have a question.<p>How likely or difficult is it for Google to engage in, for lack of better word, "thought policing"?<p>You ask your "private" AI assistant to answer a naughty question or help with problematic task(from Google's hidden list) and then you eventually face the ban hammer.<p>Did anybody ever get banned for searching the wrong keywords?
If Google is smart they'd ban Gemini access while leaving services like Gmail enabled because otherwise customers wouldn't trust them and would avoid Gemini.<p>I don't think there's any reports of banning from all Google services based on Gemini use.
> Did anybody ever get banned for searching the wrong keywords?<p>No, but they probably pass clusters of (perceived to be) dangerous searches on to the Feds. Talking out my ass though.
Solution: Use Google Drive to backup a VeraCrypt volume?
Which product do you recommend? OneDrive? Dropbox?
Filen is quite good, is E2E encrypted and currently offering (final round of) lifetime plans for Black Friday.<p>They are not super mature yet (though have been around for several years) so the product still has some improvements to be made, but I like it.
I have to imagine they are all on the lookout for CSAM. They’d simply have to be.<p>If it goes beyond that then let me know.
They're all the same to restic.
[dead]
Through my work I have access to Google's, Anthropic's, and OpenAI's products, and I agree with you, I barely touch OpenAI's models/products for some reason even though I have total freedom to choose.
If we stop for a while and really consider the value of AI tools, then comparing them on price doesn't make much sense. Any of these tools give hundreds, thousands, or tens of thousands of dollars of value per month to the user. With that in consideration they should mostly be compared on quality.
> With that in consideration they should mostly be compared on quality<p>Take a look at the comments in the thread and tell me whether there is a consensus on which AI has the best "quality". Gemini, Claude, ChatGPT are all stochastic machines; they'll give me a different output at different times for the very same query, with differences in quality each time within themselves, let alone other products.<p>I did my own checks; newer Gemini's output is consistently "good enough" for me and my family now, we individually do not use the full extent of the Pro plan (collectively, we do), and NotebookLM is something which more than one of us uses everyday; Image generation is something we use once a week or so. Given all this, the feature breadth within Gemini covers all bases for us, with a significant catch-up in quality compared to earlier to a point that we don't really need to look elsewhere for now.<p>Plus, for us USD 20 is not a small amount; it's equivalent to one of our larger utility bills we need to pay for every month. So price is definitely an important point of consideration.
The same thing is true for a _ton_ of tech products. My home internet plan easily gives me more than $1000 in value per month. My cell phone hardware probably gives me $2000+ in value over even a short 2 year life. Customers still tend to choose the cheapest option that meets requirements.
I don’t know, I ditched my ISP of many years as soon as a better option came up, even though it cost more, because it is much higher quality.
Home internet and cell phones are fungible. AI is not.<p>If Internet would suddenly become $10k a month, maybe you would change country, or move to an office.<p>If AI would suddenly become $10k you can't do anything about it.
If AI suddenly became $10k/month or even $1k/month, I would stop using it. It just doesn't provide that much value to me. If it did, I would probably find a way to use local models or some other approach to drive the cost down.<p>If home internet became $1k/month, I would pay it. $10k, no - I just don't have the cashflow to support that.<p>If I had to choose one of the three to give up, AI, home internet, or cellphone, I would give up AI. If I had to choose two, I'd give up my cell plan. Home internet is worth a ton of value and dollars to me.
It's all about the chip economics. I don't know how the _manufacturing cost_ of Google's TPUs compares to Nvidia's GPUs, for inference of equivalent token throughput.<p>But at the moment Nvidia's 75-80% gross margin is slowly killing its customers like OpenAI. Eventually Nvidia will drop its margins, because non-0 profit from OpenAI is better than the 0 it'll be if OpenAI doesn't survive. Will be interesting to see if, say, 1/3 the chip cost would make OpenAI gross margin profitable... numbers bandied in this thread of $20B revenue with $115B cost imply they need 1/6 the chip cost, but I doubt those numbers are right (hard to get accurate $ numbers for a private company for the benefit of us arm-chair commenters).
Yes, from the first principles perspective this AI thingy is just about running electricity through some wires printed on silicon by a Taiwanese company using a Dutch machine. Which means, up until the Taiwanese you have plenty of room to cut margins up until that point the costs are mostly greed based. That is Nvidia is asking for the highest price the customer can pay and they have quite a way to the cost that define their min price. Which means AI companies can actually keep getting better deals until the devices delivered to them are priced close to TSMCs bulk wafer printing prices.
I remember, maybe 2-3 years ago, chuckling at Google with their Bard naming and being late to the game and so on. It seems like I was very wrong and that they caught up quickly enough. I was also wrong in thinking MS doing well, when their recent Copilot moves across Office, Windows, and GitHub have been a joke.
IMHO Gemini surpassed ChatGPT by quite a bit - I switched. Gemini is faster, the thinking mode gives me reliably better answers and it has a more "business like" conversation attitude which is refreshing in comparison to the over-the-top informal ChatGPT default.
I've found Gemini 3.0 Pro to be bad at multi turn conversation and instruction following. It ignores your follow up question unless you draw attention to it with caps or something.<p>Not a major complaint for technical work where you don't even want to do much multi turn conversation. Just an observation.
Is there a replacement for ChatGPT projects in Gemini yet?<p>That's the only ChatGPT feature keeping me from moving to Gemini. Specifically, the ability to upload files and automatically make them available as context for a prompt.
Looks like it is coming out soon:<p><a href="https://www.androidauthority.com/google-gemini-projects-2-3620950/" rel="nofollow">https://www.androidauthority.com/google-gemini-projects-2-36...</a>
Isn't that already nouswise? Nouswise would ground answers based on low level quotes plus you get an api from your projects.
Isn’t that what gems are?
> [Gemini] has a more "business like" conversation attitude which is refreshing in comparison to the over-the-top informal ChatGPT default.<p>Maybe "business like" for Americans. In most of the world we don't spend quite so much effort glazing one another in the workplace. "That's an incredibly insightful question and really gets to the heart of the matter". No it isn't. I was shocked they didn't fix this behavior in v3.
Not quite - I'm German :P<p>But as a sibling has said, the "super nice question homie" texts are not coming (as much) in Gemini as in ChatGPT (for me). I know that you can tune ChatGPTs persona, but that changed also the answer quality for me for the worse.
> Maybe "business like" for Americans. In most of the world we don't spend quite so much effort glazing one another in the workplace. "That's an incredibly insightful question and really gets to the heart of the matter". No it isn't. I was shocked they didn't fix this behavior in v3.<p>I presume rejecting the glazing is exactly the behavior they're praising Google for. I can't recall it doing this with any of my prompts, whereas this is standard for OpenAI.
I'm a daily user of Gemini. I get this glazing every single time. This is my very last interaction with Gemini (edited for brevity),<p>> I have a young cryptomeria japonica that is about 1 meter tall, growing in the ground. Is it too late to bonsai this plant?<p>> That's an excellent question! [etc...]<p>> I have been told cutting back to brown wood will prevent back budding<p>> That is a great clarification and you are touching on a crucial point in bonsai technique! [etc...]<p>Every. Single. Time.
I get:<p>> It is absolutely not too late to bonsai your Cryptomeria japonica. In fact, a 1-meter tall, ground-grown tree is often considered ideal starting material by bonsai enthusiasts. [...]<p>And when followed up with 'I have been told cutting back to brown wood will prevent back budding' I get:<p>> That is a very common piece of advice in bonsai, but for Cryptomeria (Japanese Cedar), it is a half-truth that requires clarification. [...]<p>That's in 'Thinking with 3 Pro' mode. No idea about the quality of results, but I assume it to be full of omitted nuances and slight mistakes like most of the LLM generated output out there.<p>Maybe they tune their models to be less glaze'y for Germany? Or The Machine has Learned that you respond more positively to glazing? :)<p>I rarely use LLMs because I don't want my brain to atrophy, but when I do I use Gemini precisely because it doesn't try to tell me I'm a very smart boy.
I tried it with Gemini 2.5 Pro. I got:<p>"Excellent question!"<p>and<p>"That is an excellent and very important question."<p>I primarily use Gemini 2.5 Pro for AI coding, and it does this to me with virtually <i>every</i> prompt.<p>"That's an insightful point!"<p>"Excellent question!"<p>And on and on. I'm not exaggerating when I say it does this almost <i>every</i> time. Easily over 90% of the responses.
What helped me to get rid of such nonsense in ChatGPT is to make a custom instruction (personalization, customization) in the settings.<p>><i>Be efficient and blunt. Tell it like it is; don't sugar-coat responses. Get right to the point. Be innovative and think outside the box. Give options, explain reasoning. Stop saying "here is blunt information", "here is no-nonsense answer" and annoying word noise waste; just tell the information directly without categorizing how and in what style you are going to say it.</i>
Gemini 3 doesn't though, which was the point. If you compare gemini 2.5 then its not Googles best model.
You know you can control that, right? I'm constantly blown away by the number of posts in threads like this from people who clearly aren't aware of custom instructions.<p>Go to 'Personal Context' on the user menu and enter something like this:<p><i>Answer concisely by default, and more extensively when necessary. Avoid rhetorical flourishes, bonhomie, and cliches. Take a forward-thinking view. Be mildly positive and encouraging, but never sycophantic or cloying. Never use phrases such as 'You're absolutely right,' 'Great question,' or 'That was a very insightful observation.' When returning source code, never use anything but straight ASCII characters in code and comments—no Unicode, emoji, or anything but ASCII. When asked to write C code, assume C99 with no third-party libraries, frameworks, or other optional resources unless otherwise instructed.</i><p>ChatGPT and Claude have similar features. Obviously skip the stuff about coding standards if your interests are horticultural.<p>It will still occasionally glaze you, but not to an insufferable extent, as happens by default.
Ironically, the thing that annoys me most about Gemini is the Discord-esque loading messages in the CLI. Twee is one thing: mixing twee with serious hints is worse.
I think we are finally seeing the effects of the steady stream of departures of top researchers and leaders from OpenAI since last year. Sure you can declare a "code red", but who is going to lead the effort? Set the direction? Do the heavy lifting? Chart the path forward? Sam Altman is a salesman, not a researcher. Ilya is no longer around. Most of the other top brass has been poached by Google/Meta/Anthropic or left to start their own thing. The people left behind are probably good at iterating, but can they really make the next leap forward on their own?
> Altman said the company will be delaying initiatives like ads, shopping and health agents, and a personal assistant, Pulse, to focus on improving ChatGPT<p>It's so telling that they're delaying these "festures" because the know full well people don't want them.
The primary reason I have switched is that creative writing has plummeted on ChatGPT. It is overly eager to censor output that isn't adult but might vaguely be adult if taken incorrectly. This severely limits creative freedom. On the other hand, Gemini happily writes my stories.<p>I am not sure who OpenAI aims to please by nerfing their own product in this way. It can't be paying customers.
"Code red" feels like theater. Competition is healthy - Google's compute advantage was always going to matter once they got serious. The real question isn't who's ahead this quarter, but whether anyone can maintain a moat when the underlying tech is rapidly commoditizing.
It was always clear that the insane technological monopoly of Google would always eventually allow them to surpass OpenAI once they stopped messing around and built a real product. It seems this is that moment. There is no healthy competition here because the two are not even remotely on the same footing.<p>"Code red" sounds about right. I don't see any way they can catch up. Their engineers at the moment (since many of the good researchers left) are not good enough to overcome the tech advantage. The piling debts of OpenAI just make it all worse.
"Who is ahead this quarter" is pretty much all that the market and finance types care about. Maybe "who will be ahead next year" as a stretch. Nobody looks beyond a few quarters. Given how heavily AI is currently driven by (and driving!) the investment space, it's not surprising that they'll find themselves yanked around by extremely short term thinking.
Yeah, but now it's questionable whether the insane investments will ever pay off.
It feels like (to me) that Google's TPU advantage (speculation is Meta is buying a bunch) will be one of the last things to be commoditized, which gives them a larger moat. Normal chips are hard enough to come by for this stuff.
Also, they have all the infra to actually use all that tpus advantage (as well as actual researchers, contrariwise to OpenAI)
That will be less of a problem since OAI can spill out to other providers as needed if their own capacity is under high utilization. They already use coreweave, aws, azure, etc. Google doesn't do that as far as I know and don't see why they would, so they are stuck eating the capacity planning.
OAI is already working on shipping their own chips.
Declaring a “code red” seems to be a direct result of strong competition?<p>Sure, from an outsider’s perspective, competition is fine.
I see google partnering with different companies to mine their data for AI, but I don't see that with OpenAI. They had a good thing going with Microsoft but it looks like that relationship is a bit sour now?<p>Surely they know that they can't just keep scraping the internet to train models.<p>If I don't use a Microsoft product, I'd have to go out of my way to use an OpenAI service. But they don't have a specialized "service" (like anthropic and developers) either. Gemini is there by default with Google/Reddit. To retain their first-to-market advantage, they'd need to be the default in more places, or invest in models and services that cater to very specific audiences.<p>I think their best best is to partner with different entities. But they lost reddit and twitter, and FB is doing their own thing too, so who's left? linkedin? school systems (but ChromeBook has them beat there), perhaps telecoms preloading chatgpt apps into phones?<p>In my layperson's opinion, I think they have an access problem. Windows 11/Copilot (Github and in windows) seems to be the main access stream and people hate both, and they don't have branding there either, just back-end. There is no device you can buy, service you can get that has an OpenAI branded thing on it as a value added feature.<p>I'm sure they'll do ok, but i keep hearing they need to do a lot more than just 'ok'.
No, I don't think they'll be okay. A long slow death perhaps, but I would be surprised if they can dig themselves out of this hole.<p>You can't beat Google on high-quality data for pretraining; at scale, that's what really matters most, both in theory and practice. Other companies like Anthropic and DeepSeek are keeping up by taking advantage of smarter RL approaches, but I just don't see anyone at OpenAI with the research credentials to do that kind of work as they all left in the last mass exodus. They have been too complacent and let much of their high-quality talent go to their competition.
Since the release of Google Gemini 3 two weeks ago, the seven-day moving average of ChatGPT's daily unique active users has declined by 6%.<p><a href="https://www.moomoo.com/news/post/62341840/why-has-openai-initiated-a-red-alert-should-nvidia-also" rel="nofollow">https://www.moomoo.com/news/post/62341840/why-has-openai-ini...</a>
Is it really a race? It feels more like a slog. I continue to try to use AI (google, openai, and anthropic), and it continues to be a pain in the ass. Their consumer interfaces are garbage, both being buggy/bloated and clunky to work over multiple threads, with its "memory" being nearly nonexistent outside a single thread. They randomly fail to do the thing they did successfully 5 minutes ago. I struggle to get them to do basic things while other things they do effortlessly. They're bad at logic, spatial reasoning/engineering, and I have to constantly correct them. Often they'll do things in agents that I never asked them to do, and I have to then undo it... The time I used to spend doing things manually, I now spend in fixing the thing that's supposed to be automating the manual work... and no matter how I try to fix it, it finds a new way to randomly fail. I am <i>much happier</i> just doing things by hand.
It sounds like you have found an approach that works for you, and that's great. In my experience I've had to devote a lot of time to learning to use AI tools. Most of this learning is understanding how to create the necessary context for success and getting an intuition for what questions to ask.
OpenAI was founded to hedge against Google dominating AI and with it the future. It makes me sad how that was lost for pipe dreams (AGI) and terrible leadership.<p>I fear a Google dystopia. I hope DeepSeek or somebody else will counter-balance their power.
That goal has wildly succeeded -- there are now several well financed companies competing against Google.<p>The goal was supposed to be an ethical competitor as implied by the word "Open" in their name. When Meta and the Chinese are the most ethical of the competitors, you know we're in a bad spot...
I said DeepSeek because they are very open (not just weights). A young company and very much unlike Chinese Big Tech and American Big Tech.
Without having followed the issue of "AI Ethics" that closely, Anthropic seems to me to be relatively non-evil, too.
> Anthropic seems to me to be relatively non-evil, too.<p>Eh... maybe? We don't yet know the results, but they have been proponents of heavy regulatory interventions since forever. Their plan was basically regulatory capture, where they sell their FUD regarding alignment, "safety" and all that jazz. If they succeed that will be evil, IMO.<p>The best thing that can happen for us regular users is both healthy competition at the SotA level (which we kinda have, with the big4 labs keeping eachother honest) <i>and</i> support for small open source local models (gemmas, llamas, mistrals, qwens, etc).
AGI was the thing from the start. From the OpenAI Charter:<p>>OpenAI’s mission is to ensure that artificial general intelligence (AGI) ... benefits all of humanity.<p>I agree with you on the leadership.
Doesn’t it seem likely that it all depends on who produces the next AIAYN? Things go one way if it’s an academic, and another way if it’s somebody’s trade secret.
Crazy how we went from google feeling like they were a dinasour who could never catch up to openai, to almost feeling like the opposite in terms of being able to catch up. All within just 1-2 years.
Thats like innovators dillema in action. Google had one of the strongest ML teams years before majoriry of AI companies was founded, but no desire to make a product that will compete with their search.<p>And now they actually have competitors.
Google (generalist/media) > Anthropic (code) > x.AI (excellent price/quality balance).<p>ChatGPT is a bit late now (even behind DeepSeek with DeepThink I believe)
Personally I find the current Google products mediocre almost on all aspects. The killer feature of chat bots is voice chat and ChatGPT works great, and Gemini is extremely quiet without a way to increase volume. It's also difficult to figure out how to sign up for Gemini, or even the keyboard that I'm typing is making so many incorrect predictions.
I just don't trust Google. To me they're pure marketing and their engineering excellence ended a few years ago.
Google literally publish the attention paper. Have people not been paying attention? Google has been the only company I’ve been watching that really understands what they are doing.
I never understood this line of reasoning. I found it much more impressive that OpenAI's ML researchers realized this is the thing and bet big on it first, than to come up with it in the first place. It's underappreciated how much talent and insight it takes to see the obvious.
The TPU architecture is the most impressive thing I care about. They developed them and have been using them internally for years. This shows they grok what they're actually doing.<p>There are serious philosophical problems with betting big on an interesting outcome like ChatGPT, even though it <i>seems</i> obvious (Google also did this of course), but creating the best architecture to do that job seems like a first-principles intelligent move, because there was no reason to keep using graphics cards except that they "did the job."
The company didn't publish the paper, employees did. And all of them have since moved on to other companies, including OpenAI.
IMO Google struggles to productize things, so they sit on great ideas a while or do the wrong thing with them, but OpenAI really showed the way and Google can probably take it from here.
Google has great technology, their ability to make and focus on great product development without getting distracted is the issue
If Google wasn’t threatened by OpenAI et al., it wouldn’t be making Gemini today though.
I have (rather, had) a paid subscription to ChatGPT. I work at my home in the Sierra foothills, and on alternate weeks in my office in San Jose.<p>Last month, I used ChatGPT while in SJ. I needed a function that's only available to paying customers, and which had worked well from my home. ChatGPT refused to recognize me as a paid-up customer. I had correct login creds + ancillary identifying info, but no go. Over the course of about half an hour, ChatGPT told me in several different ways it wouldn't (not couldn't) attempt to verify my customer status.<p>I'm now a former ChatGPT customer.
Weird. I’ve traveled across Europe and used ChatGPT paid account from my phone and my laptop in multiple countries on various connections. Mobile data, home WiFi, hotel WiFi, coffee shops, etc. I always get an email to confirm the login with a code but they’ve never denied my login or prevented me from using my account thankfully.
I would be surprised if bad customer experience handling were the reason OpenAI loses to Google. It's not like Google is known for their customer experience.
Of course Google is mature enough that this particular failure mode probably won’t happen, but there may be other more concerning failure modes for individuals who are reliant on a broad swath of Google services.<p>Diversity of tech companies is an important consideration for me, one that definitely outweighs one-time issues, especially in a field where credible competition is limited.
How do you handle family obligations and a super commute like that?
My commute is every other week, so it's not terrible. I drive to SJ Sunday night, stay in a hotel that's 5 minutes from my office, then drive home Friday afternoon.<p>It averages 3.25hrs one way, or about 13 hrs/month, given my every other week schedule. It's a little tiring, but doable.
Super commuting is a thing since this whole RTO shit show happened. A lot of companies use it as excuse to lay-off.<p>As someone who does it, it depends on motivations. If the paycheck you bring in with the commute is more than what you’ll make by getting a new job, your kids are semi independent, your partner can hold the fort down Monday to Friday it’s doable.<p>It sucks but it’s doable
I mean, cool story bro.<p>So you experienced a bug, which happens on software. I've traveled a lot and have never had an issue with my ChatGPT subscription. I'm not doubting you, but I don't think your anecdote adds much to the conversation of OpenAI vs Google.
[flagged]
I declared 'code red' at my house as Google, OpenAI and Anthropic catch up in my software development career race.
They don't have much to worry about as long as Google keeps focusing on the models and neglecting the experience of actually using them.
Most comments here seem to discuss coding results. I know these are compared against industry benchmarks, but does anyone have experience using these with non CS related tasks? For example the other day I was brainstorming a kayak trip with both ChatGPT and Gemini 3.0. ChatGPT was off the rails. Trying to convince me the river flowed a different sirection than it does, and all sorts of weirdness. Gemini didn't provide information nearly as well as a human with experience, but it wasn't _useless_ information. The OpenAI model was a catasrophe at this. I'd be curious how the different models rate for the general audience, and if that plays into it at all.
This "all hands on deck" thing is a classic tactic managers use when they don't actually know what to do or have the domain expertise to allocate resources intelligently and help their employees do their jobs.
why couldn't GPT5.1 improve itself? Last I heard, it can produce original math and has phd level intelligence.
This will keep going around the table, next it might be a Chinese company that demos 98% of the capability at 1/4 the price. The objective of being at the cutting edge of LLM performance seems like more of a marketing advantage in the game of sucking in more capital for a moatless technology.
"We’re currently experiencing issues" <a href="https://status.openai.com/" rel="nofollow">https://status.openai.com/</a>
What are devs using to run Gemini agents in vscode? 2.5pro on Cline/Roo was pretty buggy compared to Claude/gpt4/5 (also using Cline /roo), kept getting stuck in loops outputting repeated text and many editing issues, and much much worse than Claude code or codex. Has it gotten better? Is there a better way of using Gemini in vscode?
But hey they dumped $6.4 billion on Jony Ive. Surely he'll solve all their problems.
This sounds like the wrong move- focusing on the product layer and counter positioning on ads is the way to beat G
Listen, I just had to go through numerous prompt cycles to 'prove' to 5.1 that we had a new Pope. ChatGPT was dead set that I was reading 'unreliable sources'. The data is _old_.
Most discussion focused on capabilities. But I wonder does OpenAI's "make a even big and costly model" strategy even work in long term? They are already losing money at current size. Unless we have some break though in chip efficiency.(which didn't seem to be likely for now) They are only going to loss even more.
Fix: Bring back Ilya, fire Sam Altman.
How have OpenAI only <i>just</i> realized this?
Is anyone actually getting good results out of GPT Pro? For coding problems, GPT Thinking seems faster and more accurate. Pro has given me some very dumb answers actually, totally misunderstanding the question. Once I asked it do design a reverse osmosis system for our home, and it suggested a 7k system that can produce 400 liters per minute. Even though I explicitly told it that a couple liters per minute suffice.
I work with Gemini 3 daily, and I think the hype is unwarranted. It takes shortcuts, hallucinates and its UI seems way behind. And what's with the small fonts?
How does Anthropic fit into this? It's much smaller but feels like they have a much clearer product definition with their Claude Code.
As for my use cases, google and especially anthropic are not "catching up". They are better for long time already
OpenAI is toast. Google has a model advantage, hardware advantage (TPUs), and business advantage (I hear they are good at selling ads).<p>It is all physics from here.
ChatGPT seems like a huge distraction for OpenAI if their goal is transformative AI<p>IMO: the largest value creation from AGI won’t come from building a better shopping or travel assistant. The real pot of gold is in workflow / labor automation but obviously they can’t admit that openly.
In one of the Indian movies, there is a rather funny line that goes like this "tu jiss school se padh kar aaya hai mein uss school ka headmaster hoon". It would translate like this "The school from which you studied and came? I am the principal of that school". Looks like Google is about to show who the true principal is
I think the most relevant quote is from Futurama:<p>"Eh-de-de-de-de. Don't quote me regulations... I co-chaired the committee that reviewed the recommendation to revise the color of the book that regulation is in. We kept it gray."
Probably all of the ML foundation like transformers which OpenAI used to create its chatbot was originally developed at Google.
AI creates the possibility to disrupt existing power structures - this is the only reason it gathers so much focus. If it were merely tool that increased efficiency of work, few would care so much. We already frequently get such tools which draw far less attention.
What will it do to Jony Ive’s legacy if his OpenAI device is no more successful than Snapchat’s foray into hardware?<p>If OpenAI becomes an also-ran by the time the hardware is released, this seems like a real possibility no matter how well-designed it is.
> What will it do to Jony Ive’s legacy if his OpenAI device is no more successful than Snapchat’s foray into hardware?<p>Well, in my opinion his legacy is already pretty tarnished by his last few years at Apple, his Love From company, and his partnership with OpenAI. If he somehow knocks it out of the park with OpenAI (something I don’t think will happen nor do I want it to) then maybe he can redeem himself a little bit but, again IMHO, he is already about as low as he can go. Whatever respect I had left for him vanished after the OpenAI/IO announcement video.
Not sure what you mean. His legacy to date is ruining the iphone because he couldn’t think of anything to do beyond “thinner”.
Why doesnt he ask chat gpt to solve it all? He sells it saying it does everything!
If OpenAI is smart here, they would figure out that you can make more money on a flop than with a hit. I bet an AI would figure that out.
This sounds like their medicine might be worse than what they're currently doing...
it's hard to get invested into anything google when they've been non stop killing products or making them worse for over a decade.
Code red?<p>Altman should know better. This sends terrible signals to employees, stakeholders and customers.<p>You don’t solve quality problems by scrambling teams and increasing pressure.<p>This reeks of terrible management. I can imagine Stanford graduates grinding it past midnight for “the mission”. If any if you is reading this: don’t do it. Altman is screwing you over. There are plenty of other places that won’t code-red your christmas season while having hundreds of billions of dollars in cash.
isn't MSFT the one screwed here. Who is on the line to provide more compute for them .
History doesn't always repeat... but it sure as hell rhymes.
When I was playing poker for living there was a spreadsheet meme. There was always some guy who was losing consistently but declared everything will change from tomorrow because he now made a spreadsheet with an exact plan going forward.
The spreadsheet usually contained general things like 8 hours of sleep, healthy food, "be disciplined", "study the game for 2 hours a day" etc.<p>Of course it never worked because if he knew what he should be doing he would be doing it already instead of hoping for spreadsheet magic to change the course.<p>>>There will be a daily call for those tasked with improving the chatbot, the memo said, and Altman encouraged temporary team transfers to speed up development.<p>Sam Altman clearly didn't get the memo.
For once, capitalism works
The fate of OpenAI is effectively sealed - it will go bankrupt and the scraps will get absorbed by Microsoft, for further enshitification. Not necessarily the "end" of AI, but enjoy your account while it's useful.<p>The problem is, there is a whole ecosystem of businesses operating as OpenAI API wrappers, and those are gonna get screeeeewed.
This is the system working.<p>Competition is all you need.
Source: <a href="https://www.wsj.com/tech/ai/openais-altman-declares-code-red-to-improve-chatgpt-as-google-threatens-ai-lead-7faf5ea6" rel="nofollow">https://www.wsj.com/tech/ai/openais-altman-declares-code-red...</a> (<a href="https://news.ycombinator.com/item?id=46118396">https://news.ycombinator.com/item?id=46118396</a>)
It’s funny because it wasn’t long ago Open Ai was telling everyone else it’s game over.
Going short OPAI.PVT 10x leverage.
I’ve preferred Claude over ChatGPT for over a year so not sure what he’s on about.
I have the research to win the race. These people are masters of the fog.
Related:<p><i>TPUs vs. GPUs and why Google is positioned to win AI race in the long term</i><p><a href="https://news.ycombinator.com/item?id=46069048">https://news.ycombinator.com/item?id=46069048</a><p><i>Google, Nvidia, and OpenAI</i><p><a href="https://news.ycombinator.com/item?id=46108437">https://news.ycombinator.com/item?id=46108437</a>
Does anyone have a link to the contents of the memo?
OpenAI fragmented into multiple companies that are now competing against them. OpenAI is buying compute and data.<p>Meanwhile, Google consolidated their AI operations under Google Deepmind and doubled down on TPUs.<p>The strategy "solve AGI and then solve everything else" is an all-in gamble that somehow AGI is within reach. This is not true.
Google fragmented into multiple competing companies as well, that's where OpenAI itself came from. The problem is even after shedding employees into all these startups or established competitors trying to catch up, Google has way more people, money, and compute to throw at things and see what works than the rest of the industry. It's demoralizing and tempting for people to go back, which is also demoralizing
> There will be a daily call for those tasked with improving the chatbot, the memo said, and Altman encouraged temporary team transfers to speed up development.<p>It's incredible how 50 year-old advice from The Mythical Man-Month are still not being heed. Throw in a knee-jerk solution of "daily call" (sound familiar?) for those involved while they are wading knee-deep through work and you have a perfect storm of terrible working conditions. My money is Google, who in my opinion have not only caught up, but surpassed OpenAI with their latest iteration of their AI offerings.
Besides, can't they just allocate more ChatGPT instances to accelerating their development?
> It's incredible how 50 year-old advice from The Mythical Man-Month are still not being heed.<p>A lot of advice is that way, which is why it is advice. If following it were easy everyone would just do it all the time, but if it's hard or there are temptations in the other direction, it has to be <i>endlessly</i> repeated.<p>Plus, there are always those special-snowflake guys who are "that's good advice for you, <i>but for me</i> it's different!"<p>Also it wouldn't surprise me if Sam Altman's talents aren't in management or successfully running a large organization, but in machiavellian manipulation and maneuvering.
Imho it just shows how relatively simple this technology really is, and nobody will have a moat. The bubble will pop.
Not exactly. Infra will win the race. In this aspect, Google is miles ahead of the competition. Their DC solutions scale very well. Their only risk is that the hardware and low level software stack is EXTREMELY custom. They don't even fully leverage OCP. Having said that, this has never been a major problem for Google over their 20+ years of moving away from OTS parts.
amazing how the bubble pops either from the technology either being too simple or being too complex to make a profit
the thought that this might be done one recommendation of ChatGPT has me rolling<p>think about it, with how much bad advice is out there in certain topics it's guaranteed that ChatGPT will promote common bad advice in many cases
Also, google has plenty of (unmatched?) proprietary data and their own money tree to fuel the money furnace.
There is always a daily call if a U.S. startup fails. Soon there will be quadrants and Ikigai Venn diagrams on the internal Slack.
Don't forget the bleak subtext of all this.<p>All these engineers working 70 hour weeks for world class sociopaths in some sort of fucked up space race to create a <i>technology that is supposed to make all of them unemployed</i>.
These engineers make enough money to comfortably retire by the time they are replaced with AI.
> technology that is supposed to make all of them unemployed.<p>To make all of <i>us</i> (other poor fuckers) unemployed.
You can have a more upbeat take on it all.
They are paid exceptionally well though. Way above market rate for their skill set was at any point in history.
Work long hours for a few years and enjoy freedom for the rest of your life. That's a deal a lot of people would take. No need to feel sorry for the ones in position to actually get the choice.
Wait, shouldn't their internal agents be able to do all this work by now?
<p><pre><code> There will be a daily call for those tasked
with improving the chatbot, the memo said,
and Altman encouraged temporary team transfers
to speed up development.
</code></pre>
Truly brilliant software development management going on here. Daily update meetings and temporary staff transfers. Well known strategies for increasing velocity!
Don't forget scuttling all the projects the staff has been working overtime to complete so that they can focus on "make it better!" <i>waves hands frantically</i>
"The results of this quarter were already baked in a couple of quarters ago"<p>- Jeff Bezos<p>Quite right tbh.
…someone even wrote a book about this. Something about “mythical men”… :D
I've had ideas for how to improve all the different chatbots for like 3 years, nobodys has implemented any of them (usually my ideas get implemented in software somehow the devs read my mind, but AI seems to be stuck with the same UI for LLMs), none of these AI shops are ran by people with vision it feels like. Everyone's just remaking a slightly better version of SmarterChild.
I really want a UI that visualises branching. I would like to branch out of specific parts of the responses and continue the conversation there but also keep the original conversation. This seems to be a very standard feature but no one has developed it.
Oh man, I hadn’t thought of SmarterChild in dog’s years! It was an early AIM chatbot, and felt like magic at the time. Looking back it feels like there’s a clear through-line from it (and the rest of ActiveBuddy’s menagerie) to the ChatGPTs of the world today…<p>For today’s lucky 10,000, here’s a Vice retrospective from 2016:<p><a href="https://www.vice.com/en/article/a-history-of-smarterchild/" rel="nofollow">https://www.vice.com/en/article/a-history-of-smarterchild/</a>
Did you open-source / publish these ideas?
i agree - it shows a remarkable lack of creativity that we're still stuck with a fairly subpar UX for interacting with these tools
Its easy to dismiss it but what would you do instead?
What if they make 2 daily calls, that would surely improve the velocity by 2 times!
<i>The beatings will continue until morale^H^H^H^H^H^H chatGPT improves...</i>
what do you mean "catches up"<p>Gemini has been as good as GPT for more than a year<p>OpenAI still somehow gets the edge on the initial veneer of hype, and that's running thin
I think most people are aligned on AI being in a bubble right now with the disagreement being over which companies (if any) will weather the storm through the burst and come out profitable on the far side.<p>OpenAI, imo, is absolutely going to crash and burn - it has absolutely underwhelming revenue and model performance compared to others and has made astronomical expenditure commitments. It's very possible that a government bailout partially covers those debts but the chance of the company surviving the burst when it has dug such a deep hole seems slim to none.<p>I am genuinely surprised that generally fiscally conservative and grounded people like Jensen are still accepting any of that crash risk.
To be honest, this is the first month in almost a year when I didn't pay for ChatGPT Pro and instead went for Gemini Ultra. It's still not there for programming, where I use Claude Max, but for my 'daily driver' (count this, advice on that, 'is this cancer or just a headache' kind of thing), Gemini has finally surpassed ChatGPT for me. And I used to consider it to be the worst of the bunch.<p>I used to consider Gemini the worst of the bunch, it constantly refused to help me in the past, but not only has it improved, ChatGPT seems to have gone down the 'nerfing' road where it now flat out refuses to do what I ask it to do quite often.
Google is too big to fail. It's the backbone of the Internet. Just YouTube is synonymous with online video.
A hardware device from OpenAI is exactly why I would prefer it over Anthropic or Google. Why give up on differentiation? I would assume the model team is separate from the consumer hardware team.
OpenAI was founded a non-profit to benefit humanity. Why does the "race" matter?
OpenAI is for-profit: <a href="https://www.theguardian.com/technology/2025/oct/28/openai-for-profit-restructuring" rel="nofollow">https://www.theguardian.com/technology/2025/oct/28/openai-fo...</a>
They gave up on that a long time ago.
We are in a pretty amazing situation. If you're willing to go down 10% in benchmark scores, you easily 25% your costs. Now with Deepseek 3.2 another shot across the bow.<p>But if the ML, if SOTA intelligence becomes basically a price war, won't that mean that Google (and OpenAI and Microsoft and any other big model) lose big? Especially Google, as the margin even Google cloud (famously a lot lower than Google's other businesses) requires to survive has got to be sizeable.
Google trains its own AI with TPU's, which are designed in house. Google doesn't have to pay retail rates for Nvidia GPUs, like other hyperscalers in the AI rat race. Therefore, Google trains its AI for cheaper than everyone else. I think everyone else "loses big" other than Google.
Surely they can just use AI to go faster and attend their daily calls for them...
Conspiracy time.<p>>be Google<p>>watch regulators circle like vultures<p>>realize antitrust heat is rising faster than stock buybacks can hide<p>>notice a small lab called OpenAI making exotic tech and attracting political fascination<p>>calculate that nothing freezes regulators like an unpredictable new frontier<p>>decide to treat OpenAI as an accidental firebreak<p>>let them sprint ahead unchecked
watch lawmakers panic about hypothetical robot uprisings instead of market concentration<p>>antitrust hearings shift from “break up the giants” to “what is AGI and should we fear it”<p>>Google emerges looking ancient, harmless, almost quaint<p>>pressure dissipates<p>>execute phase two: acceleration
roll out model updates in compressed cycles<p>>flood the web with AI-powered services<p>>redefine “the internet” as “whatever Google’s infrastructure indexes”<p>>regulators exhausted from chasing OpenAI’s shadow<p>>Google walks back onto the throne, not by hiding power, but by reframing it as inevitability
conspiracy theorists argue whether this was 5D chess or simple opportunism<p>>Google search trends spike for “how did this happen”<p>>the answer sits in plain sight:<p>>attention is all you need
It's a fun idea but there's ample public reporting about how Google reacted to the rise of ChatGPT. There is reporting that Google was taken by surprise. You can be skeptical of that, but that's what the reporting says. ChatGPT went viral in Nov/Dec 2022, and by February or March Google was scrambling to stand up Bard as a viable competitor.<p><a href="https://web.archive.org/web/20221221100606/https://www.nytimes.com/2022/12/21/technology/ai-chatgpt-google-search.html" rel="nofollow">https://web.archive.org/web/20221221100606/https://www.nytim...</a><p><a href="https://web.archive.org/web/20230512133437/https://www.theverge.com/2023/5/12/23721037/google-ai-progress-search-docs-starline-video-calls" rel="nofollow">https://web.archive.org/web/20230512133437/https://www.theve...</a>
That would be believable if you forget the sheer incompetence and bureaucracy Google was/is filled with
there is enough proof that they had a chatbot internally which was quite competitive but was not pushed through for all these fears, it seems they were always confident that they could catch up and scaling laws were their internal defense.<p>The question now though is neither might have expected Chinese labs to catch up so fast.
This is one conspiracy theory I've actually considered. Google waited until the Chrome outcome to come out swinging.
I take this code red as a red flag. Open AI should continue to concern itself with where it will be 5 years from now, not lose sight over concern about where it will 5 months from now.
open ai is at risk of complete collapse if it cannot fulfill its financial obligations. if people willing to give them money don't have faith in their ability to win the AI race anymore, then they're going out of business.
Spoiler alert they're going to go out of business
Exactly. They aren't going to win the AI race chasing rabbits at the expense of long-term goals. We're 3 years into a 10 year build-out. Open AI and it's financiers are too impatient, clearly, and they're fucking themselves. Open AI doesn't need to double it's revenue to meet expectations. They need to 50x their revenue to meet expectations. That's not the kind of problem you solve by working through the weekend.
The financiers are running out of money to lend. At this point, staying negative profit isn’t an option, they need to be able to fund themselves or they’ll go bankrupt.
i cannot imagine how they are going to be able to meet their obligations unless they pull off a massive hail mary at this point via a bail out or finding someone to provide tens of billions of dollars in funding.
Back in the day before Adobe bought Macromedia, there was a constant back and forth between Illustrator and Freehand where each release would better the competitor at least until the competitor's next release.<p>Does anyone in AI think about 5 years from now?
You can't make a baby in 1 month with 9 women, Sam.
[dead]
[dead]
Google is shivering! /s
It’s actually code yellow
Word needs need OpenAI and Anthropic like startups to drive AI forward. Think about only Google, Meta, MS, AWS is only have these capabilities. They will never able to do that in one hand, other hand it will be monopolistics. We need more AI startups, not monopolies.