Several current and former Openai researchers talk about the company’s first social media invasion: the SORA application, a Tiktok flow filled with AI videos and many Deepfakes Sam Altman. The researchers, who emit their complaints on X, seem to be torn on how the launch fits Openai’s non -profit mission for the development of advanced and benefits humanity.
“AI -based feeds are scary,” Openai Pretraining John Hallman said in a Post in x. “I will not deny that I felt some concern when I first learned that we released Sora 2.
Boaz Barak, another Openai researcher and Harvard professor, responded: “I share a similar blend of concern and excitement. Sora 2 is technically amazing, but it is premature to congratulate ourselves on avoiding the traps of other social media applications.”
Former Openai Rohan Pandy researcher used the moment Connect a new bootMagazine workshops, consisting of former AI Lab researchers trying to build AI systems for scientific discovery: “If you do not want to build the infinite AI Tiktok machine, but you want to develop AI that accelerates fundamental science … Come with us to workshops.”
There was many other Publications along the same lines.
Sora’s launch highlights a basic tension for Openai that launches again and again. It is the fastest growing consumer technology company on Earth, but it is also an AI Frontier workshop with a high non -profit map. Some former Openai employees I talked about to argue that the consumer business can, in theory, serve the mission: Chatgpt helps finance AI research and distributes technology widely.
Openai Sam Altman CEO said so much to a Post in x Wednesday, addressing why the company has so many funds and computing power in a social media app:
TechCrunch event
Francisco
|
27-29 October 2025
‘We mainly need capital for construction [sic] AI who can do science, and we are definitely focused on AGI with almost all our research effort, “Altman said.” It is also nice to show people cool new techniques/products along the way, make them smile and hope to make some money for all these needle computers. “
“When we started chatgpt, there were many” who needed this and where is Agi, “Altman continued.”[R]Eality is thin when it comes to optimal orbits for a company. ”
But at what point does Openai’s consumer business exceed its non -profit mission? In other words, when does Openai say no to an opportunity they develop money because it is in contrast to the mission?
The question emerges as regulators, examining the transition of Openai to the speculative, which OpenAi must complete to collect additional capital and ultimately to be made public. California Attorney General Rob Bonta said Last month that “it is particularly concerned with securing the reported Openai security mission as a non -profit nature” remains ahead and center “in restructuring.
The skeptics rejected Openai’s mission as a brand tool to entice the talent from the Big Tech. But many internalists in Openai insist that they are central to why they joined the company in the first place.
At the moment, Sora’s footprint is small. The app is a day. But his debut marks an important expansion of Openai consumer businesses and exposes the company to motivation that have contaminated social media applications for decades.
Unlike Chatgpt, which is optimized for utility, Openai says Sora is built for fun – a place to create and share AI clips. The supply feels closer to the Tiktok or Instagram wheels, platforms that are famous for their addictive loops.
Openai insists he wants to avoid these traps, supporting in a blog Announcement by Sora’s launch that “concerns about Doomscrolling, Addiction, Insolation and RL-Sloptimized foods are top minds”. The company explicitly says it is not optimized for the time spent on feed and instead wants to maximize creation. Openai says it will send reminders to users when they have been rolling for a long time and will mainly show them people they know.
This is a stronger starting point than Meta’s vibes-another short circuit video released by AI released last week-which appears to have struggled without so much safeguards. As a former leader of OpenAi policy, Miles Brundage point out It is likely that there will be good and bad applications of Ai-Video supplies, as we have seen in the Chatbot era.
Still, as Altman has long recognized, no one wants to create an addictive application. The motives for the operation of a driver lead them to it. Openai has still faced problems around Sycophance in Chatgpt, which the company says it was unintentional because of some of its training techniques.
To one June PodcastAltman discussed what he calls “the great bad alignment of social media”.
‘One of the big mistakes of the social media era was [that] The power algorithms had a bunch of unintentional, negative consequences in society as a whole and perhaps even individual users. Although they did the thing a user wanted – or someone believed that users wanted – at the moment, which is [to] Take them, as, keep spending time on the site. ”
It’s too early to say how aligned the Sora app is with users or the Openai shipment. Users are already observing some techniques that come from involvement in application, such as the dynamic emojis that appear every time you like a video. This feels designed to shoot a small dopamine to users to deal with a video.
The real test will be how Openai is evolving Sora. Given how much AI has taken over regular social media foods, it seems reasonable that Ai-Native Feeds could soon have their time. Whether Openai can grow Sora without reproducing the mistakes of its predecessors remains to be seen.
