Sora’s shutdown could be a reality check moment for AI video

OpenAI announced this week that it will shut down Sora and related video models just six months after launching the app.

In the latest episode of TechCrunch’s Equity podcast, Kirsten Korosik, Sean O’Kane, and I discuss what the decision means for OpenAI and the industry more broadly. To some extent, this move seems consistent with what we’ve heard about OpenAI as it focuses on enterprise and productivity tools ahead of a potential IPO.

In fact, Kirsten suggested that OpenAI’s decision to shut down Sora was “a sign of maturity that would have been nice to see in the AI ​​lab.”

But Sora’s shutdown — along with ByteDance’s announced delay in the worldwide launch of its Seedance 2.0 video model — could also serve as a reality check moment for makers of AI video tools, and for evangelists who claim these tools will replace Hollywood anytime soon.

Read a preview of our conversation, edited for length and clarity, below.

Anthony: I think it’s worth highlighting that it’s not just about the app. I mean, the app wasn’t attractive to me, at least, and I think to others, because it was the idea of ​​a social network without people, where it’s nothing but shit.

But beyond the app, OpenAI appears to be essentially killing everything it does with video. According to the Wall Street Journalwho broke some of this news, it really comes down to this idea that Open AI — before potentially going public — is really trying to focus on business products, enterprise products, and programming products. [So] This social app for the consumer, [and] More broadly, video, is not a priority right now.

Shawn: Yes, I haven’t really used it [the app]. I was intrigued by the idea of ​​this for a few different reasons. And, you know, that was a good reminder that Open AI – and I don’t mean to bring them down in any way – but I think that was a reminder to them, probably, internally, of the element of luck. […] How successful is ChatGPT?

There’s clearly something of value to people, and I don’t want to take that away, because you don’t get to the usage numbers that we’ve heard about without something that’s working properly – and what’s more is that it’s been maintained over a number of years and developed into something that continues to be meaningful to people.

But there was an element of Sora, when he showed up, like, “We’ve built the most successful consumer product ever, and now we’re doing it again. We’re going to bring in Disney and all this stuff.” I think this is just a really harsh reminder that it’s not always going to be an absolute shortcut to the pinnacle of the greatest consumer products of all time, and that there really has to be something that people feel like they’re getting some meaning out of for it to last.

Kirsten: Yes, I actually want to give props to OpenAI for this decision, because sometimes we scoff at the whole “move fast and break things” idea, but I think there is some value [to] Companies that can iterate very quickly and then throw out products that don’t work and never feel failure behind them. I mean there was real money lost. If you look at the deal with Disney, That was a billion dollar dealbut if you look at — and we don’t have insight into this because we don’t see their balance sheets — but what were they spending on this and what is the long-term value of the company?

And I think that, certainly, it would have been interesting to see what they could create, but their decision to shut it down, to me, showed a sign of maturity that would have been nice to see in the AI ​​lab.

Anthony: In terms of what this means for OpenAI, it seems very consistent with everything we’ve heard about their strategy moving forward. It doesn’t seem like a slam dunk or anything like that in terms of how we think about the future of generative AI.

Especially in the video, it’s interesting because it also comes at this time where there have been reports about Seedance, which is ByteDance’s generative AI model [for video]. There are reports of this [Seedance 2.0 has] It was delayed mainly due to engineering and legal issues [figuring out]”Can we build IP protection into this?” Which they didn’t seem to take seriously before.

So, it’s a reality check moment. There have been these really exaggerated statements, including from people inside Hollywood [were] Like, “We’re done, this is the future, it’s just writing prompts and making feature films.” It turns out that for all technical and legal reasons, it is not that easy and we are a long way from it happening.

Shawn: And the last thing I think we should say about this as well, is that this is one of many decisions that seem to happen after the arrival of Fidji Simo [and began] A type of day-to-day operations management. This is just a huge dynamic that has changed within OpenAI. And I think the further we get away from that moment where it was used to run the show, especially these consumer products and determine their fate, the easier it will be to look at this moment in time and think about how important that moment was to this company.