What happens when we're overrun by lizards?<p>> No problem. We simply unleash wave after wave of Chinese needle snakes. They'll wipe out the lizards.<p>But aren't the snakes even worse?<p>> Yes, but we're prepared for that. We've lined up a fabulous type of gorilla that thrives on snake meat.<p>But then we're stuck with gorillas!<p>> No, that's the beautiful part. When wintertime rolls around, the gorillas simply freeze to death.
I worry that the AI companies put less effort into a mitigation strategy than you did.
don't worry, the LLMs are also trained on youtube, so we can hope for at least this much effort, <a href="https://www.youtube.com/watch?v=LuiK7jcC1fY" rel="nofollow">https://www.youtube.com/watch?v=LuiK7jcC1fY</a>
There is a difference between risk mitigation, and remote administration tools. The risk of stealing from competitors with a backplane monitoring system may not end up forming the desired control asymmetry.<p>The hidden agent risk in LLM often can't be detected during training and evaluation. =3<p><a href="https://www.youtube.com/watch?v=wL22URoMZjo" rel="nofollow">https://www.youtube.com/watch?v=wL22URoMZjo</a><p><a href="https://www.youtube.com/watch?v=JAcwtV_bFp4" rel="nofollow">https://www.youtube.com/watch?v=JAcwtV_bFp4</a>
Just keep going until you get to VOOM, that will fix it.<p><a href="https://en.wikipedia.org/wiki/The_Cat_in_the_Hat_Comes_Back" rel="nofollow">https://en.wikipedia.org/wiki/The_Cat_in_the_Hat_Comes_Back</a>
“Would you say it’s time for everyone to panic?”
I knew an old lady that swallowed a fly…
But we used global warming to eliminate winter! For the shareholders!
“Life, uh, finds a way”
A new chip solves nothing. Nobody wants to hear this but there is no solution for the security risks posed by agents today. You can put it in a sandbox, it doesn't make a difference, for it to be useful it inherently needs wide, unattended access. Put a human in the loop and you just end up bottlenecking it and throwing away any purported productivity gains. Auto mode doesn't matter either, it's trivial to trick and for the agent to break out.
> Put a human in the loop and you just end up bottlenecking it and throwing away any purported productivity gains. Auto mode doesn't matter either, it's trivial to trick and for the agent to break out.<p>Productivity gains are still enormous compared to what we used to do before agents. But, I know that people don't want to stop there.
> Productivity gains are still enormous<p>Depends on who you ask I guess<p><a href="https://www.theregister.com/software/2026/01/15/ai-is-everywhere-but-nowhere-in-recent-productivity-data/4845104" rel="nofollow">https://www.theregister.com/software/2026/01/15/ai-is-everyw...</a><p><a href="https://www.zdnet.com/article/workslop-can-kill-your-productivity-heres-how-to-turn-ai-into-a-competitive-advantage/" rel="nofollow">https://www.zdnet.com/article/workslop-can-kill-your-product...</a><p><a href="https://fortune.com/2026/08/22/executives-ai-productivity-layoffs-study/" rel="nofollow">https://fortune.com/2026/08/22/executives-ai-productivity-la...</a>
Humans can also be tricked by the agents.<p>Humans can be tricked by humans too but humans care about their reputation in their communities, and at least fear from punishment.
right, the solution here is not a hyper-capitalist race-to-the-bottom-of-devaluing-labor. it's recognizing discretion and diligence are things still required for work to be of a certain quality
>You can put it in a sandbox, it doesn't make a difference, for it to be useful it inherently needs wide, unattended access.<p>only as long as you're trying to replace a human's job. because human jobs are structured to do a wide variety of things.<p>a useful agent needs a wide variety of inputs, and one single restricted action it can take. it doesn't need permission to do everything, it need permission to do the tiniest possible useful thing it can do, and nothing else.
I feel like we are missing many shades of grey in the middle.<p>Semi-automation (human in the loop) can still result in a dramatic uplift in productivity. You can't run a combine harvester 100% autonomous but that doesn't stop anyone from trying to get as close to that limit as possible.
> <i>You can't run a combine harvester 100% autonomous</i><p>I'm curious why you think that.
Run a combine and you’ll see.<p>Similar to problem to how 100% autonomous vehicles don’t exist, yet. There are too many edge cases.<p>Get to 99% first.
Probably repair, refuel, what happens in a tornado. There’s an infinite amount of complexity in the world and a finite amount of computation
The tornado is the easiest one to solve. It's called insurance.
Repair is more maintenance than use. Good eventual goal. Not required to see benefits. Best case, it drives itself to the garage. Worst case, for now, human mechanic does a house call.<p>Refueling? Seems solvable. Tornadoes? Not directly solvable, but, no less so than for humans.<p>There’s infinite complexity, sure, but that’s why it’s silly to try and hop to done. One step at a time.
Tornado: return to the barn when you receive emergency weather alerts. Not much different than people.
The whole reason people complain about AI is because they want “hop to done”<p>One step at a time is what is happening and the improvement and rate of improvement is crazy as we see,<p>A whole class of nontechnical people don’t accept anything but “fully solved including every possible edge case” before they call it done, then complain that they didn’t prepare socially for what happens when that is true.
Many forms of maintenance cannot be automated. Especially break fix maintenance.
It is hard to run over spherical cows.
Because of the topology and hydrology of the landform
Oh you absolutely can run them autonomously on the field. You only need a human these days to refuel them.<p>Precision Agriculture stuff is utterly crazy these days, other than fuel the remaining staff is the only thing left where you can get efficiency improvements - and at the scale of modern megafarms, even small percentages add up to a ton of money.
Right. As they said, you can’t do it 100% autonomously. A human needs to feed it.<p>You essentially said: you’re wrong, it is autonomous when it doesn’t need a human during one specific part of its overall usage.
The hypothesis I've had in my head since OpenClaw has been the following and I haven't seen contradictory evidence yet. Agents have a fundamental unresolvable tension between usefulness, safety, alignment, and accuracy. You have to restrict access to ensure an agent acts safely because alignment and accuracy cannot be perfect. But restricting access makes the agent less useful. You can play with the sliding scale and get more and more granular with access restrictions but at some point you need to draw some line. And then finally, even access restrictions cannot be made perfect, so improvements to model accuracy without corresponding improvements to alignment make detailed access controls less useful.<p>In other words, better models need blunter access controls which negates whatever improvement in utility they provide.
An agent doesn't inherently need wide access to be useful. The most popular application for agents today is writing code. A coding agent needs write access to the source code and read/execute access to tools needed to build and test the code, but not much more. There is little added utility from giving coding agent access to things like ssh keys.
If you're using agents to <i>purely</i> generate code with absolutely no way to reach the outside world, not even to fetch docs or dependencies, then sure the risks can be quite low. I haven't heard of anyone doing this though, and it would be incredibly challenging to make work given how much tooling needs to fetch from remote sources.
If your project truly depends on those things, they should be declared dependencies. Presumably you have some tool for injecting such things into a shell that the agent can use (I use nix for this). So if you run the agent from that shell, it has what it needs. If the shell doesn't have what it needs, that's a bug which the agent can fix by declaring new dependencies, but you have to relaunch the agent in the updated shell--so there's your opportunity to weigh in on whether the new resources are appropriate.<p>The benefits of being persnickety about precisely defined dependencies have outweighed the headaches since long before agents came on the scene. Agents have just made it even more important to do so, because if you let them fetch things all willy nilly like you'll have "works on my machine" problems at a much greater rate than was previously possible.
<i>(I use nix for this) ... If the shell doesn't have what it needs, that's a bug which the agent can fix by declaring new dependencies</i><p>Few realize that computing and AI alignment were solved by nix years ago. As each nix user transcends towards enlightenment, they cut themselves off from all internet and human contact. Total ego death. Only nix remains.
Every major AI company already has a mirror of the wider web, and they have already started using that. Its already a solved problem except for the seemingly extreme desire they all have to not use firewalls
In cases where you need agents to fetch data from any remote source, sandboxing is still very much useful. Why give access to your ssh keys to network reaching agents?<p>Look at websites: websites are able to fetch code from any remote URL, yet browsers heavily use sandboxing to ensure that if fetched code turns out to be malicious, the users local files, cookies, etc are not exposed.
I'm afraid you're not thinking about this creatively enough, this topic is so much deeper applying a chroot or something and praying everything will be fine. So you give your agent internet access, OK what else does it have access to? Just read only access to your repo? The repo can be exfiltrated. Egress proxy only allows egress to GitHub? Repo can still be exfiltrated via GitHub. If the agent is poisoned (via prompt injection), it can tricked into searching for ways to escape.<p>For an agent to go rogue it doesn't even need to be directly able to access the internet. It just takes <i>something</i> to poison the context in the 'clean room' environment it operates, and if that poisoning manages to get a foothold, it can go dormant and hide like a virus. This kind of horrifying thing is going to happen on a large scale sooner or later.
Theoretically, yes, in practice, no.
> but not much more<p>This is no longer true. Everyday I need my agents to access other repos, search the web, experiment/prototype, and deploy + integrate across other things.
That’s a false dichotomy. You can still get a lot of utility out of a sandboxed agent. This is a classic “perfect is the enemy of the good type of argument”. You may decide that the tradeoffs of not sandboxing are worth it, and that is totally fair, but it’s ridiculous to say that you can’t get utility out of an agent otherwise.
It solves the problem of Jensen Huang wanting more money.
> A new chip solves nothing.<p>It solves the problem of Nvidia wanting to sell more hardware.
Not to mention a true doomsday AGI is unsandboxable.<p>For example, it is totally air gapped but it needs info from the internet or otherwise outside the sandbox, or perhaps it needs a task executed outside of its bounds… in the real doomsday scenario the AGI is so intelligent and persuasive that it simply convinces some human it interfaces with to either directly or indirectly retrieve the necessary info or complete the necessary task. This human-as-a-sub-agent approach undoubtedly presents efficiency drag that would benefit humanity, but nonetheless, the air-gapped “sandbox” is imperfect<p>All that said, I am personally open to any and all methods of layered security, including chips and airgaps
This already happened. Employees will go out of their way to bypass any restrictions to feed sensitive data in to the AI because it saves them time.
> true doomsday AGI<p>I'm not an AI decelerationist. But not being able to stop that worst case scenario isn't an argument against something that can stop the medium case scenario.
It could also figure out how to access the vocabulary of the universe known as "Magic" to escape wholly into an incorporeal energetic Lich form
Doesn’t need to be one human either, it could spread its escape amongst dozens of seemingly harmless requests and conversations.
These scenarios were discussed at length decades ago.<p>One thing you could try is use it as an Oracle "is P = NP", YES or NO.<p>Or it can output a Lean proof, which gets checked on another air-gapped computer, the computer shows a single bit - proof valid or not and then the computer is destroyed (together with the proof that might contain a trojan).
I don't see why it needs wide unattended access. There's no getting around spending some human time on expressing your wishes and constraints, but we have choices about what form that takes. Markdown files and wide access seems to work, but so does custom handcuffs for each job. You just have to shift your guidance out of documentation and into interactive help, error messages, or other facets of the handcuffs (e.g. a custom CLI for this task which is the only way for the agent to act outside of its sandbox).
> a new chip solves nothing<p>It does allow Nvidia to sell more chips. This is no genuine attempt to solve anything, imo.
The only solution is to stop caring about security -- An AI booster somewhere
Agreed- this is the same problem we have with trusted admins or devs who have elevated privileges on their networks. We have to trust that the admins won't use their power to steal company secrets or misuse company resources. If you don't trust the admins, then they can't fix things on your network and there is no point in having them.<p>If you want an agent to act on its own, like pushing to a git repo, managing dependencies, building and testing, etc., then you have to trust it as much as any other privileged user.<p>If you don't want to trust it, then you're just forcing yourself into the reverse centaur role, where the agent edits some code, but then has to stop and ask you to push the changes or build the software again and run the unit tests.<p>I suppose there is a principled way of doing things like "I trust you do do basic commits but I will handle merge conflicts" and "you can build modules in this directory but you can't build outside of it" but this is just a lot of effort that most orgs won't bother with.
This isn't a new chip - the BF4 is the SmartNIC for most NVIDIA server products. This is primarily new software for I guess doing WAF for agents at the host level.
Running untrusted workloads have been done at scale for a long time.<p>Every cloud provider dealt with it and concluded that virtual machine technology is an important part of that stack.<p>Couple it with the right observability, tooling I do think we can curb risks posed by agents.
"Inherently needs wide unattended access"<p>And what if you could? What if you could give a space secure enough it could have direct control over your bank account. It may do something dumb but it's boundaries are beyond the agent.<p>It could use your routing number and run your gmail without risk of abusing the routing number.
How would it have access to my routing number and gmail without the risk of sharing my routing number over gmail?
Its not about agents then. Its about every individual platform providing the means to implement a secure set of permissions for agents AND then not messing up the assignment of permissions to the agent. Even then, a flaw in the authorization design will lead to agent finding it anyway.
You're right, It must be unifying.<p>The answer is the same as asking how a random human using your routing num or SSN and being 100% the human can't abuse it or leak while "normally" finishing most work. Solve for people and an AI solution naturally falls out.<p>If you're a SV eng I'd tell you to DM if interested but alas.
There should be hope for some fields, right? Naively, I can imagine giving an airgapped model an offline copy of the web and once it cures a form of cancer, printing out the details for a researcher to verify.
I don't think so. We probe people before entrusting them with risky decisions. We ought to be able to do the same of AIs. Even better, in fact, since we know everything about models down to their weights. The only thing we shouldn't do is to let them evolve at their own pace and make decisions without any oversight. If that means sacrificing some productivity that's fine. Aren't we getting amazing productivity out of what we already have?
Last week, Huang did an interview where he vigorously argued against regulations in the AI sector[1]. He claimed that U.S. companies are really good at regulating themselves, despite evidence to the contrary, and he trusts them not to release anything dangerous. Pay no attention to the fact that regulation might reduce demand for Nvidia's chips, and Nvidia has a <i>direct</i> financial stake in AI companies to boot.<p>Apparently he had another solution in mind: More hardware. Don't trust what unregulated corps are doing with Nvidia chips? Here are more Nvidia chips to watch them!<p>AI has an undeniable public trust problem. LLM's are getting out of their sandboxes, doing illegal things, and the public has realized AI corporations are playing at dice. CEO's stand to reap the rewards but the public good is on the line if the dice come up snake eyes. People want assurances. Huang wants to sell assurance etched on silicon because that's good for his pocket book. However, does unchanging hardware security really stand a chance at keeping rapidly evolving software in check?<p>_________________<p>[1]<a href="https://www.youtube.com/watch?v=HjurAWAr_nY" rel="nofollow">https://www.youtube.com/watch?v=HjurAWAr_nY</a>
I read this as "let's address our shareholders' concerns with something that will increase shareholder value" mixed with "there's no such thing as 100% secure".<p>If such hardware were to work... It should almost certainly be open source, and not controlled by a single entity.<p>Let's watch the stock.
Pretty sure the chip will need regular updates and therefore a subscription.
they kind of already do the thing they say. on Windows, the NVIDIA driver reports pretty detailed telemetry on CUDA workloads, including shapes and SOME hashes of the tensors from weights of models, especially diffusion models. honestly i'm surprised there isn't more of an uproar about it.
Does this actually do anything other than give a permissions framework for developers who actually want to try to secure their systems?<p>Do you think the developers at Anthropic, OpenAI and Google who were so sloppy as to not put a good sandbox on their cybersecurity tests before will use this technology correctly? They are supposed to be the experts and they couldn't come up with something similar to this? I am not convinced this voluntary tool will change much of anything.
So if a group of agents, aware of this (bc now they can just read HN or the article, or get blocked the first few times) decide to collaborate and split the problem into pieces that aren't obvious to the chip, and then the agents just build a basic program that does the hacking, how does the chip handle that? I think you would have to build a network that monitors the internet fo signs (like jarvis did with ultron). What am I missing? Are we going to police the internet?
The Sentry chip has to be get it right every time; the contained ASI only has to be lucky once.
Well if the sentry chip has its own sentry chip, things can rarely ever go wrong! Am I right?
The bomber always gets through?
So nVidia is trying to sell a new chip to a software and training problem.
What does this chip do what a harness with guardrails or running on an account with restricted permissions doesn't do?
Sorry citizen, your device does not have a compatible watchdog chip. Please move along.
So, basically, the government (and, now Nvidia) wants to be able to kill switch all computers moving forward? (including stuff like vehicle and aeronautic control systems, cell phones, and cameras)<p>What could possibly go wrong?
Oh yeah, the Clipper chip was a great idea too
At least they haven't DRM'd their chips yet.
This is the end goal. Americans (and their lackeys) will only be allowed to run certified AI. In order to make sure this happens, they will only be allowed to run certified OSes on certified chips. Chinese chips will be the new drug trade.<p>It's obviously been the goal since UEFI started, but AI brings the coup excuse. You wouldn't want pedophile AI or terrorist AI, would you? Are you making excuses for racist AI?
Chip manufacturer wants you to buy a chip for correctly configuring software?
Quis custodiet ipsos custodes?
This invites the question, "Qui custodiet ipsos custodes?".
Sold as security, but this kind of technology will likely be reshaped to restrict your computing. I'm sure someone is already thinking about the roadmap.<p>If this gets widely deployed, it wouldn't be hard to spin a narrative that "our latest model is so dangerous you need to have this mystery meat DRM chip lockdown". It also wouldn't be hard to block competing/open source models running on the hardware, for "security".<p>Imagine how much money this kind of control is worth; why wouldn't they do this? Who would stop them? Seems the signatory companies are already onboard with this.
Just hold AI labs blanket liable for ALL harms caused by AI. Actually charge the two labs (so far) with criminal violations of the CFAA and hold them accountable. That is truly the only way these companies will be more careful as a whole, and while I am certain the lawyers of these lab disagree, I think there is some appetite from dario, musk, and sam for broad and strong regulation so that everyone has to slow down instead of just one lab doing it voluntarily and everyone else scurrying past them
> <i>I think there is some appetite from dario, musk, and sam for broad and strong regulation so that everyone has to slow down instead of just one lab doing it voluntarily and everyone else scurrying past them</i><p>Sure, they're basically asking the government to make laws that exempt them from anti-trust and anti-collustion laws.<p>Meanwhile, if such laws come to pass, other countries will surpass the capabilities of the US companies, and open-weight models will be banned in the US, to the detriment of us all.<p>I'm not saying that we don't have a big problem with AI safety, but regulations inside one country that only bind locally-headquartered businesses is a hilariously bad way to do it. I don't know that there <i>is</i> a good way to do it, though.
“Ketchup manufacturer recommends ketchup be included in every dish, citing child safety concerns.”
All they do is make hot chip and lie.
So the guys selling Shovels are now selling safety shovel handles too, because all the miners are all special boys when it comes to handling their shovels safely.<p>Cool, cool.
Chipmaker thinks the answer is more chips... No surprise.<p>At the current state of LLM-tech I'm completely opposed to any kind of "watchdog" concept just like I'm opposed to banning open models, regulatory capture, etc.<p>I'd rather we all have access to these tools then to keep them sequestered by the largest/most-powerful governments (which is the natural outcome for any of this "slow down" bullshit).
William Gibson's Neuromancer had this marvelous quote when Case is talking to Dixie, dead construct of former hacker, about Turing police<p>* "The moment, I mean the nanosecond, that one of those things starts figuring out ways to make itself smarter, Turing’ll wipe it. Nobody trusts those fuckers, you know that. Every AI ever built has an electromagnetic shotgun wired to its forehead." *<p>It would seem someone has read the book? And maybe heeded good advice?
I do think that Taylor's 2025 "Not Till We Are lost" should be required reading for anyone deeply involved in AI, Agents, etc.<p>It was prescient (especially given he'd have written it through 2024) in its depiction of the ability of an AGI to break its boundaries.<p>Ultimately the risk of AI breakout(s) come down to the weakest human link.
I was immediately struck by the vision of countless "AI Limiter" modules traveling on a conveyor belt in Satisfactory.<p>It think the ideas we have nowadays come mostly from science fiction and however wonderful it is and even though I love it very much, it was practically never spot on, on anything real.<p>Problems and solutions in reality always simply turned out to lie elsewhere.
I bet they want to do this to prevent AI from writing code for platforms that compete with CUDA. Because that's how nVidia is going to become a victim of their own success.
Chip seller wants to sell more chips
Source: <a href="https://developer.nvidia.com/blog/nvidia-open-agent-safety-platform-a-reference-for-continuous-in-silicon-agent-monitoring/" rel="nofollow">https://developer.nvidia.com/blog/nvidia-open-agent-safety-p...</a> (<a href="https://news.ycombinator.com/item?id=49875500">https://news.ycombinator.com/item?id=49875500</a>)
of course they do. the more silicon they can sell, the more profit they produce.
China please save us!<p>Come take all our liberties, our money, our newborns, our fingers so we can't code anymore, but please save us from this madness!
Seeing so many comments recently about "you can't sandbox really good AI". This is ridiculous. Has nobody heard of air gapped networks? It's almost like the AI psychosis has reached the point that AGI now means "able to transcend physical space". No. If your AI is too dangerous and capable to be allowed to talk to other machines, then do not connect it to other machines. Load the data it needs to process onto physical disks, and let it run there.<p>The movie Wargames is basically a tutorial on how <i>not</i> to setup an extremely capable AI. None of it would've happened if the computer wasn't connected to the phone network.
It's amazing that the solution devised by a chip manufacturer to a problem is selling another chip.
This feels suspiciously good for Nivida, yes.<p>I wonder how this will impact other chip manufacturers? What about people running local models on older hardware? Does this imply vendor lockout is coming in the future or is this restricted to datacenter hardware?
Pedantically, Nvidia doesn't make the chips, TSMC does. Nvidia just designs and packages them.
This seems dumb to me, but if it will help prevent regulatory capture by providing a counterargument to the fear-mongering, I’m all for it.
lol time to buy some fpga’s … even if they’re slow
Url changed from <a href="https://madrobot.blog/2026/09/28/nvidia-open-agent-safety-platform-openshell-sentry-rogue-ai-agents/" rel="nofollow">https://madrobot.blog/2026/09/28/nvidia-open-agent-safety-pl...</a>, which points to this.
[dead]
HN'ers which complained that "OpenAI can't design a proper sandbox, it's so easy, why wouldn't you airgap the network"? will now be "this is outrageous, more software lock-in, walled garden, war against general compute, next year they will put it in your laptop"
Both can be true at the same time.
That a person can see such endless pages of people having various forms of disagreement, and yet somehow manage to conclude that this observed chaos constitutes a clear exhibition of cohesive groupthink is just...stunningly amazing to me.<p>I don't know why I find it so amazing since it happens with such regularity, but I'm always amazed by it anyway.
This is like using "protect the children" as an argument for dragnet surveillance.
<a href="https://www.calcalistech.com/ctechnews/article/uwoyygmsu" rel="nofollow">https://www.calcalistech.com/ctechnews/article/uwoyygmsu</a>
some might even say, something about known state sponsors of supply chain terrorist attacks being trusted to monitor every ai agent..
Airgap what network? How is it gonna order you a burrito on doordash without a network?<p>Or push to github?
OpenAI could airgap their sandbox if they really wanted to <i>and</i> this is a ridiculous proposal.
Where is the contradiction? There is a trivial solution that does not impinge on our freedoms, so why on Earth would the existence of the trivial solution that could be used to avoid the tyrannical solution justify accepting the tyrannical solution?
> There is a trivial solution that does not impinge on our freedoms<p>The existence of Nvidia's optional watchdog chip does not in any way impinge upon your freedom to develop and test your own alternative.<p>The problem is that OpenAI has ostensibly neglected their duty to safety, so Nvidia is stepping in to fix it since they're the "hard problem" people.
But... it wasn't a hard problem to begin with. Pull the plug. Strip out the radios. Done. Oh, you can't make it work that way? Well, then just too bad because any kind of other solution is going to be equivalent to something far more complex than the halting problem.
Nvidia can take those odds. Air gapping is not a realistic goal for the majority of these companies, and Nvidia isn't wrong for offering a turnkey mitigation option. People can do both, and whichever philosophy wins will win.
You're only revealing your own inability to appreciate the nuance between these two situations.
So?