ChatGPT unveils Sora with up to 20-second AI video generation OpenAI has been promising to release its next-gen video generator model, Sora, since February. On Monday, the company finally dropped a working version of it as part of its “12 Days of OpenAI” event. “This is a critical part of our AGI roadmap,” OpenAI CEO Sam Altman said during the company’s live stream. According to the OpenAI team, Sora will be made available to Plus and Pro subscribers in the U.S. and around the world starting Monday afternoon. YouTuber Marquis Brownlee reportedly got early access to the video generator and released a brief review on his channel on Monday morning. Sora appears to not be built atop GPT-4, as virtually all of OpenAI’s other generative tools are. The model is not available through the standard ChatGPT website, but instead through Sora.com (which is still not live as of this post’s publication). The model is capable of generating videos in resolutions ranging from 480p to 1080p in lengths from 5 to 20 seconds, from either text prompts or reference images. It’s also capable of editing and extending existing video clips. ChatGPT Plus subscribers will be allowed up to 50 clip generations at up to 720p per month, and fewer videos at higher resolutions, each five seconds long. Pro users will be allowed unlimited generations at all resolutions and durations as long as 20 seconds. In addition to editing tools, Sora also offers a “storyboard” feature that will enable creators to combine multiple prompts into a single cinematic scene. Related One of ChatGPT’s latest features comes to the free tier Researchers call ChatGPT Search answers ‘confidently wrong’ The best AI tools of 2024: all the generative AI apps you need to try Brownlee notes that the model needs “a few minutes” to generate a 1080p clip, but notes “that’s also, like, right now, when almost no one else is using it. I kind of wonder how much longer it’ll take when this is just open for anyone to use.” Brownlee also points out that the model has significant difficulty in properly generating legs and their movements, with the front and rear legs swapping positions in unnatural and incomprehensible ways. Unlike Grok 2, Sora will limit what its users can create and explicitly bars the generation of copyrighted subjects, people under the age of 18, and anything containing violence or “explicit themes.” Despite OpenAI’s leading position in the AI industry, Sora has been beset by delays throughout its development, enabling competitors like Runway’s Gen-3 alpha, Kuaishou Technology’s Kling and Meta’s Movie Gen models to beat it to market. Sora was also recently (however briefly) publicly leaked by a group of beta testers, who accused the company of “art washing” the model’s capabilities.
I got Fictional: 1. Atticus Finch 2. Dr. Ellie Sattler Historical 1. Rachel Carson 2. Benjamin Franklin ChatGPT just trying to butter us all up.
Or maybe it knows the complexities of my wife all too well and is like dude.... it's a lose lose situation. You can't win this war. Get a GC and cut call it a day.
I don’t think Hollywood workers need to worry quite yet. T-3000: https://bsky.app/profile/labuzamovies.bsky.social/post/3ld2c4wls322j
I really don't know what all of this means but im sure its pretty big. I'd use ChatGPT to summarize but chatgpt has become so nerfed that I would probably walk away with more questions than answers. And for whatever reason, ChatGPT and Grok sound just alike these days. I heard microsoft is out.
Lena Khans corpse is still warm and Trump has already created the next MIC company ready to abuse government tax dollars.
Microsoft isn't "out" since Microsoft owns almost 50% of OpenAI from what I recall. Stargate was announced about a year ago, I think, and Microsoft/OpenAI were already discussing the buildout of the infrastructure like the data center(s) back then. I'm guessing, like the tweet said, Azure is heavily involved (?), although Oracle has their own offering, as well. The articles I read about it last year had hypothesized this thing would need a small sun (aka, a nuclear reactor) to power it, but I haven't heard anything about that since then. Maybe since Katy has evacuated, they could plop a nuclear reactor there and use it.
Its all confusing to me. OpenAI still has major issues with being nonprofit. And Elon's stinky finger is somewhere in this mess. If anything, it reminds me of the Artemis program. A giant road to nowhere.
All the og members of openai are fleeing from their ceo with suitcases full of options while hyping up for-profit startups with insanely unproven valuations. The contract they have with MS says something like the deal being off when they hit AGI, but no one wants AGI without profit, so they redefined it as something that generates a hundred billion plus dollars in profits, prob because Elon forced the issue, Anyhow, Stargate is interesting in the idea that the government is propping up the industry from an "ai winter" even while it's insanely rare to scale big profits from ai similar to early dotcom companies. It could loosen up the tech jerbs market, but not many are qualified to become ai specialists let alone roughly 1/5 of out of the 100k. And if Stargate succeeds, who knows what the overall jerbs market will be. I guess its good for the government to take that seriously... Frankly, the government is better off with the Deepseek Jina model of reverse engineering private company discoveries and working off open source like llama where Zuckerberg is burning billions in hopes of making a platform standard. The ROI is 10-100x cheaper. Instead, this helps Zuckerberg and others shave off some research costs while other countries don't really need to take the lead and spend hundreds of billions in investment. I don't hate Stargate, but a half a trillion in spending is always a big deal. Optimistically, it could be an Al Gore invents the internet kind of moment, though adding bigger advances into ai research has been mostly limited to adding the most computing power...
None of it makes sense. It reeks of a giant grifting scheme. OpenAI goes public, gets a massive valuation. OpenAI then proceeds to dump excessive amounts of money into NVidia, raising NVidia's valuation even further. Why does Oracle even exist? Softbank loves to invest in hot new companies with no real growth value for the long term - Uber, Weworks, Doordash, Bytedance, Sprint, Boston Dynamics, Cruise, OneWeb ... all of these companies will be a shell in 10 years. Meanwhile, Softbank capitalizes on its NVidia investment. Its puzzling Musk is quiet about this. Maybe its an opportunity to have 'competition'. IMO, it seems Anthropic and xAI seem to have more potential.
Out of mild curiosity, I returned to my ChatGPT account and gave it a simple book request (recommend a few reliable books on topic X). I know about this topic so would not need to do a lot of extra research to check its work. And the very first one was a hallucination. It gave the wrong author married to a title of an article (and not a book). Hilariously, it gave a two sentence summary of the "book" that matched neither the listed author's work, nor the article with that title. The algorithm just hemmed around, word by word, as it does. Naively, it seems trivial to fix this for references, with just 1-2 extra lines of code from Open AI. But I'm sure I'm oversimplifying. I do tell my students or any interested scientist two points of caution: * hallucinations will continue on referencing (not that big of a deal -- just don't ask these bots to do certain things) * careful what chunks of you own work you ask a bot to edit. If you have any IP, you may have just offered it up, for free, to an algorithm or god knows who. Some universities offer contracts with grammarly, for instance, to their students. While it's a good tool, I think, grammarly's privacy statements make it very plain that it's sucking up content and doing with it what it pleases. Maybe not a big deal if you're just proof-reading a bland essay as a student, but maybe less good if you're working with something creative, exploratory, or entrepreneurial.
The free version of ChatGPT has become unusable. The $20.00 version is ok, but I still had a lot of problems with it, at least to the point where it was not worth the $20.00. It seems ChatGPT is greatly reduced the amount of tokens allowed per request.
1. Point it at a problem, including problematic hiring or challenging human factors 2. Pew-pew! 3. ??? 4.. Great success!
