4 comments

  • lxe31 minutes ago
    If you haven&#x27;t been following the current explosion of video gen models, including all the content you see on social media, you should pay attention.<p>I have no doubt in my mind that right now, in August of 2026, given a couple of weeks, an amateur with a high-end last-gen GPU or somebody with maybe a thousand bucks worth of AI credits, given a good amount of exploration&#x2F;tinkering to fill the gaps that GenAI cannot currently do, (which is storyboarding, screenwriting, characters, and all the actual creative stuff), can crank out a high quality 1h+ feature film with only minor uncanny artifacts.<p>The reason why this isn&#x27;t happening en masse right now is that most creatives are probably not focused on generating feature films, but rather content that&#x27;s more digestible by social media users.<p>I would expect to start seeing a LOT of development in the upcoming months as some form of commodification where generated content can be both high quality and tailored to the user.
    • Planktonne25 minutes ago
      &gt; I have no doubt in my mind that right now [...] an amateur with a high-end last-gen GPU or somebody with maybe a thousand bucks worth of AI credits [...] can crank out a high quality 1h+ feature film with only minor uncanny artifacts.<p>Then why are there no examples of this? There aren&#x27;t even examples of steps towards it--the half hour film that&#x27;s nearly production quality, etc.
      • garethsprice0 minutes ago
        Zack London&#x27;s recent cyberpunk short &quot;The Patchwright&quot; (21 minutes)[1] feels like a glimpse into the AI feature film future. I like that rather than trying to replicate a traditional film it leans into the maximalist AI aesthetic.<p>1: <a href="https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=-Rzl7nUdEs4" rel="nofollow">https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=-Rzl7nUdEs4</a>
      • coffeefirst18 minutes ago
        Or even 30 second ads that don’t look like shit.<p>The technology is interesting as an element within the wider process of special effects, animation, and video games. The idea that you’re actually making a movie with the current products—and I’m a former indie filmmaker, I’d have a blast with this if it were real—is a fantasy.
      • deltaqueue17 minutes ago
        There are. FX artists are already coming to grips with this: <a href="https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=BRESQ8NX-us" rel="nofollow">https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=BRESQ8NX-us</a><p>&quot;yadda yadda, this is the worst it will ever be&quot;
      • unit-vector12 minutes ago
        I think this comes pretty close <a href="https:&#x2F;&#x2F;higgsfield.ai&#x2F;@higgsfield.studio&#x2F;projects&#x2F;oneiric" rel="nofollow">https:&#x2F;&#x2F;higgsfield.ai&#x2F;@higgsfield.studio&#x2F;projects&#x2F;oneiric</a>
      • echelon3 minutes ago
        Very few of these people are doing this on home rigs. That&#x27;s way too slow for getting any real work done.<p>I work in this industry. I&#x27;ve dealt with major studios and up-and-coming studios quite extensively. (I was a filmmaker before AI, which helps a lot.)<p>If your exposure to AI video is ComfyUI, that&#x27;s the consumer &#x2F; hobbyist segment and you&#x27;re missing out on where all the action is actually concentrating.<p>There are five-ish segments:<p>- Large studios. They&#x27;re going slow, but they&#x27;ve already started integrating the tech. They won&#x27;t tell people they&#x27;re using it. They outsource to production houses that use it. Some of the biggest studios are moving slowly and want fully air-gapped support that runs on their existing cloud contracts.<p>- Up-and-coming new media studios. Check out Gossip Goblin [1], [2], [3]. These folks are getting millions of dollars of funding right now and they&#x27;re getting huge followings.<p>- Marketing, B2B: people are already using this in ads. See Coca-Cola<p>- Consumer, UGC, non-creatives: this is the Sora crowd that only knows how to remix popular IP. Some of them graduate into new media exploration, but this follows the 1% rule. This is where most &quot;slop&quot; comes from.<p>- Porn creators: I&#x27;ve interviewed several folks that are making mid-six figures on &quot;fan&quot; platforms. It&#x27;s a full-time job and some of them are scaling up to teams.<p>[1] Pomegranate, one of the best AI films: <a href="https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=fyZhC2TXgcs" rel="nofollow">https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=fyZhC2TXgcs</a><p>[2] Theatrical release: <a href="https:&#x2F;&#x2F;variety.com&#x2F;2026&#x2F;film&#x2F;news&#x2F;gossip-goblin-ai-film-gods-dont-give-gifts-theaters-1236818068&#x2F;" rel="nofollow">https:&#x2F;&#x2F;variety.com&#x2F;2026&#x2F;film&#x2F;news&#x2F;gossip-goblin-ai-film-god...</a><p>[3] <a href="https:&#x2F;&#x2F;www.gossipgoblin.studio&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.gossipgoblin.studio&#x2F;</a>
    • an0malous7 minutes ago
      Then why isn’t there even one good, 2-minute video produced with AI yet? Can you share one?<p>A few months ago, Coca Cola couldn’t even make a good 1 minute commercial with millions of dollars and a whole team of video producers and editors. It ended up just being a montage of 5 second clips of animals smiling at Coca Cola trucks.
    • bri3d25 minutes ago
      There’s a technology fundamental driving the short form aspect too, current video models are all trained on &lt;500 frame sequences and struggle to generate long shots. There are a lot of attempts in the literature to improve cross-generation coherence but there needs to be an architectural breakthrough or significant scaling improvement before longer continuous shots are practical. So you’d be making a really ADD feature film at the moment.
    • mrdootdoot10 minutes ago
      No job is forever. Tomorrow isn’t promised, and automation already did this—albeit more slowly—throughout the second half of the 20th century. Some people watched their jobs disappear and accepted that they’d become obsolete. Plenty of others pivoted.<p>The counterargument today is that we can pivot much faster. I’m typing this on a $2,000 Mac that can run Blender, render scenes, write code, edit video, and access tools that would have required a studio or a room full of specialized hardware not that long ago.<p>GenAI is empowering people to do things they simply could not do before. It’s here. The interesting question is: what are you going to do with it?<p>I think movie production may be approaching its YouTube moment. If the barriers to making something collapse, then make something worth watching. I’d rather see a small team with an actual idea than another $300 million spider man multiverse hot garbage sequel or guardians of the slop 600 assembled by committee, approved by a board after a slide deck, and polished into expensive slop.
      • jasonjayr5 minutes ago
        The <i>really</i> interesting question is: What are we going to do when a technology so quickly and thoroughly displaces the need for human labor, in a society that constantly demands you trade your labor for the right to exist and participate in society?
    • unixhero27 minutes ago
      All I want is Seinfeld Forever
    • add-sub-mul-div16 minutes ago
      If I&#x27;m understanding correctly. Replacing low budget advertising slop is the priority over replacing $100m movie budgets simply because they haven&#x27;t gotten around to the latter yet. The capabilities are definitely there for the latter, just no one&#x27;s bothering. Okay.
  • borzi49 minutes ago
    No they are not. They are helping rip off delusional investors who think this will amount to more than slop.
    • pllbnk1 minute ago
      They are only talking about production, productivity, quantity, nothing else. The amount of produced content hasn&#x27;t been in short supply. There is no talk about consumption. Who is going to consume all this content, _even_ if it somehow turns out good? If it&#x27;s cheap and easily producible, then it&#x27;s worthless as consumption item.
  • mistrial91 hour ago
    Southern California Edison utility district famously required their office staff to train offshored support staff, and then did replace those office staff with the newly trained offshore workers<p>Southern California Edison IT workers ‘beyond furious’ over H-1B replacements - Computerworld 2015 -- The incident sparked legislative interest in California, with lawmakers proposing bills to restrict utilities from outsourcing critical infrastructure work and prompting at least 11 U.S. senators to request federal investigations into the use of H-1B visas in such outsourcing deals.
    • teddyh40 minutes ago
      Relevant: &lt;<a href="https:&#x2F;&#x2F;www.threepanelsoul.com&#x2F;comic&#x2F;on-workforces" rel="nofollow">https:&#x2F;&#x2F;www.threepanelsoul.com&#x2F;comic&#x2F;on-workforces</a>&gt;
      • stego-tech16 minutes ago
        TPS hits the nail on the head beautifully, here.<p>Look, the genie is out of the bottle: we are a global economy, and increasingly a global society. Everywhere on the planet is reachable via data link within seconds, and ~95% physically within a day - provided money is no object. Every single country is going to have to grapple with this immediately, not at some point in the future.<p>On the one hand, businesses will always want to have the lowest possible costs. On the other, consumers frequently want local service and support, especially when things go wrong. Countries and Governments want to drive business growth to increase their tax revenue and improve the quality of life of their citizens, but offshoring jobs lowers tax receipts and increases the number of citizens on social services. AI throws gasoline onto this bonfire, demanding immediate action rather than continued can-kicking. The net result is immediate pain as every country struggles to find a path forward.<p>The countries who succeed will be those who find a balance between increasing tax revenue to fund social welfare programs that their populace requires for necessities (housing, food, education, healthcare), while also promoting business growth in a global marketplace, but also ensuring labor is kept proportionate to the local customer market size. The US is trying to turn back a clock to a time of national markets, while China and the EU increasingly acknowledge the global reality and regulate or subsidize accordingly. Businesses clearly want the freedom to compete globally, but a regulation and governance scheme that’s not as fragmented as the present landscape.<p>TL;DR: outsourcing to AI or offshoring is a last-century solution for National companies to increase margins. The name of the game going forward is distributing labor and supply chains relative to local market sizes to build a balanced global competitor.
  • ck21 hour ago
    What I am curious to learn is how many episodes of a TV series made by real actors it will take for an &quot;AI&quot; to eventually make perfect extensions<p>ie. feed it every episode of the original Star Trek series, at some point &quot;AI&quot; is going to be advanced enough to write a new episode and simulate all the characters perfectly<p>anime and animation in theory will be even easier to do perfectly?<p>the question is can it do it perfectly with just a dozen episodes? or does three dozen make it better?<p>because once that happens, that&#x27;s all the networks will pay for, after all they are only about the commercials they can insert into content regardless of source<p>and yes viewers will eventually accept it the same way they won&#x27;t quit X or meta, etc. regardless of how horrible they become or the sub-human behavior of the owners
    • WarmWash1 hour ago
      Why wouldn&#x27;t the users just generate it themselves?
      • homarp44 minutes ago
        on their own computer using open weight chinese models?<p>More seriously<p>1) &gt;users will generate their own<p>some people write fanfic, most just watch&#x2F;read instead.<p>2) <a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Rebel_Without_a_Crew" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Rebel_Without_a_Crew</a> explains how film maker will film, with whatever they have. AI will enable some people to make what is not financially feasible right now for a single person to do.<p>&#x27;Unique&#x27; movie&#x2F;stories creation will continue.<p>And like with the rest of things AI can do, the question is more &#x27;who&#x27;s going to pay for person-made vs &quot;algorithmically optimized AI generated story&#x27;.<p>Same question we have of &#x27;AI generated code&#x27; vs &#x27;this code was written by a person&#x27;.
        • ygxhbcxdg36 minutes ago
          Sora was QUICKLY shut down due to the astronomical cost. On what basis do you claim that using language models to generate feature films would be cheap?
          • xhkkffbf19 minutes ago
            Sora isn&#x27;t cheap, but regular films regularly cost more than $100m to film. Spending $20k on tokens is relatively cheap.
            • lotsofpulp14 minutes ago
              I would be surprised if production costs for films regularly exceeds $100M.<p>I assume the vast majority of 9 figure films have marketing as the 9 figure expense, including pay for high profile actors&#x2F;directors&#x2F;producers, which is also a form of marketing.
      • mawadev56 minutes ago
        This is exactly how I see it. Most companies participating in such games are digging their own grave long term, because of fear of someone else doing that
      • ygxhbcxdg36 minutes ago
        because it&#x27;s too expensive. Go try out Sora to see for yourself. oh right you can&#x27;t
      • hackable_sand38 minutes ago
        Why would they?
    • WhyIsItAlwaysHN57 minutes ago
      I don&#x27;t think it will ever be able to make perfect extensions, because there is no perfect extension. Part of creative work is human imperfections. Part of me being invested in a series is because I know it will end at some point.<p>If there&#x27;s infinite star trek episodes, I don&#x27;t care to watch it. I&#x27;d rather spin a fidget spinner instead.
      • ck244 minutes ago
        you are reacting as if the content is going to be robotic<p>except &quot;AI&quot; doesn&#x27;t have any original thoughts of its own, everything is going to be originally from a human source<p>it might be a concept from some rare book they fed it and destroyed after, it might be a concept from thousands of &quot;fan fic&quot; they hoovered up all over the internet by scraping forums illegally<p>you say you won&#x27;t watch it except human behavior already demonstrates there are millions of other people who will<p>it&#x27;s an unfortunate future but an easily predictable one<p>all about making cheap content to insert ads, the ads are the point, not the content<p>like product placement in a $100 million action movie, except soon they will be able to make the movie for $100 instead of $100 million and most won&#x27;t know the difference
        • qsera17 minutes ago
          &gt; most won&#x27;t know the difference<p>First, prove it in porn. Then let us talk.
        • dan_ggggg39 minutes ago
          I&#x27;m sorry but it sounds like you don&#x27;t have a very good understanding of what art is or what role it plays in human existence. What kind of grades did you get in English class as a kid?
        • hwhshsh39 minutes ago
          no it will be uncanny slop just like all the AI slop I&#x27;ve already had forced on me<p>&gt; there are millions of other people who will<p>yeah millions of people like all kinds of dreck. sure, you probably liked The Big Bang Theory but some of us have taste
          • dan_ggggg37 minutes ago
            I&#x27;m not a fan, but it&#x27;s a massive insult to compare a show written by a team of writers like The Big Bang Theory to AI generated TV slop for illiterate retards.