Sam Altman: OpenAI has been on the ‘wrong side of history’ concerning open source

To cap off a day of product releases, OpenAI researchers, engineers, and executives, together with OpenAI CEO Sam Altman, answered questions in a wide-ranging Reddit AMA on Friday.

OpenAI finds itself in a little bit of a precarious place. It’s battling the notion that it’s ceding floor within the AI race to Chinese language firms like DeepSeek, which OpenAI alleges may’ve stolen its IP. The ChatGPT maker has been attempting to shore up its relationship with Washington and concurrently pursue an formidable knowledge middle venture, whereas reportedly laying groundwork for one of many largest financing rounds in historical past.

Altman admitted that DeepSeek has lessened OpenAI’s lead in AI, and he stated he believes OpenAI has been “on the improper aspect of historical past” with regards to open sourcing its applied sciences. Whereas OpenAI has open sourced fashions previously, the corporate has typically favored a proprietary, closed supply growth strategy.

“[I personally think we need to] work out a unique open supply technique,” Altman stated. “Not everybody at OpenAI shares this view, and it’s additionally not our present highest precedence … We’ll produce higher fashions [going forward], however we’ll preserve much less of a lead than we did in earlier years.”

In a follow-up reply, Kevin Weil, OpenAI’s chief product officer, stated that OpenAI is contemplating open sourcing older fashions that aren’t state-of-the-art anymore. “We’ll positively take into consideration doing extra of this,” he stated, with out going into higher element.

Past prompting OpenAI to rethink its launch philosophy, Altman stated that DeepSeek has pushed the corporate to probably reveal extra about how its so-called reasoning fashions, just like the o3-mini mannequin launched as we speak, present their “thought course of.” Presently, OpenAI’s fashions conceal their reasoning, a technique supposed to forestall rivals from scraping coaching knowledge for their very own fashions. In distinction, DeepSeek’s reasoning mannequin, R1, reveals its full chain of thought.

“We’re engaged on exhibiting a bunch greater than we present as we speak — [showing the model thought process] can be very very quickly,” Weil added. “TBD on all — exhibiting all chain of thought results in aggressive distillation, however we additionally know folks (at the least energy customers) need it, so we’ll discover the fitting method to steadiness it.”

Altman and Weil tried to dispel rumors that ChatGPT, the chatbot platform via which OpenAI launches lots of its fashions, would enhance in value sooner or later. Altman stated that he’d prefer to make ChatGPT “cheaper” over time, if possible.

Altman beforehand stated that OpenAI was dropping cash on its priciest ChatGPT plan, ChatGPT Professional, which prices $200 per thirty days.

In a considerably associated thread, Weil stated that OpenAI continues to see proof that extra compute energy results in “higher” and extra performant fashions. That’s largely what’s necessitating initiatives equivalent to Stargate, OpenAI’s lately introduced huge knowledge middle venture, Weil stated. Serving a rising consumer base is fueling compute demand inside OpenAI as properly, he continued.

Requested about recursive self-improvement that is likely to be enabled by these highly effective fashions, Altman stated he thinks a “quick takeoff” is extra believable than he as soon as believed. Recursive self-improvement is a course of the place an AI system may enhance its personal intelligence and capabilities with out human enter.

In fact, it’s price noting that Altman is infamous for overpromising. It wasn’t way back that he lowered OpenAI’s bar for AGI.

One Reddit consumer requested whether or not OpenAI’s fashions, self-improving or not, could be used to develop harmful weapons — particularly nuclear weapons. This week, OpenAI introduced a partnership with the U.S. authorities to present its fashions to the U.S. Nationwide Laboratories partly for nuclear protection analysis.

Weil stated he trusted the federal government.

“I’ve gotten to know these scientists and they’re AI consultants along with world class researchers,” he stated. “They perceive the ability and the boundaries of the fashions, and I don’t assume there’s any probability they only YOLO some mannequin output right into a nuclear calculation. They’re good and evidence-based and so they do a number of experimentation and knowledge work to validate all their work.”

The OpenAI workforce was requested a number of questions of a extra technical nature, like when OpenAI’s subsequent reasoning mannequin, o3, can be launched (“quite a lot of weeks, lower than a number of months,” Altman stated); when the corporate’s subsequent flagship “non-reasoning” mannequin, GPT-5, may land (“don’t have a timeline but,” stated Altman); and when OpenAI may unveil a successor to DALL-E 3, the corporate’s image-generating mannequin. DALL-E 3, which was launched round two years in the past, has gotten relatively lengthy within the tooth. Picture-generation tech has improved by leaps and bounds since DALL-E 3’s debut, and the mannequin is not aggressive on quite a few benchmark assessments.

“Sure! We’re engaged on it,” Weil stated of a DALL-E 3 follow-up. “And I believe it’s going to be definitely worth the wait.”