- While OpenAI is currently valued at $157 billion, many co-founders, leading researchers, and scientists have left the company.
- Many former employees say that OpenAI is focused more on shipping profit-oriented products over prioritizing safety research.
- OpenAI CEO Sam Altman reportedly asked investors to not back rival AI startups, in a bid to stifle competition.
While OpenAI CEO Sam Altman has secured multi-billion-dollar investments from tech giants, the internal turmoil at OpenAI has once again taken center stage due to multiple exits at the top tier. Last month, OpenAI CTO Mira Murati resigned from the AI company. Along with her, two leading researchers, Bob McGrew and Barret Zoph, left OpenAI.
Bob McGrew was the Chief Research Officer at OpenAI and Barret Zoph was the Vice President of Research. The news broke out amid Reuters reporting that the ChatGPT maker is planning to do away with its non-profit board and become a for-profit corporation.
In the last five months, OpenAI has seen high-profile exits including Ilya Sutskever, John Schulman, Jan Leike, and others. It must be noted that last year, Sam Altman was ousted by the board, but he returned just after five days. Since then, things have not been normal at OpenAI.
A recent report by The Wall Street Journal sheds light on what happened behind the scenes, and the ensuing internal conflict that led to the departures of as many as 20 researchers, top scientists, co-founders, and executives.
OpenAI Deviates From Its True Mission
The WSJ report quotes several former and current OpenAI employees who say that OpenAI is no longer a research-based organization. Under Altman’s leadership, the focus has now shifted to shipping moneymaking products as soon as possible. Employees agree that it’s run more like a business operation rather than an AI lab.
OpenAI is no longer a research-based organization, say former and current employees.
The report says that OpenAI is unable to maintain the balance between a research-focused lab and a growing business. Former AI scientists who worked at OpenAI went on to say that “massive profits have corrupted OpenAI’s culture.”
The report further adds that researchers were given only nine days to perform safety evaluations on the GPT-4o model. Researchers worked 20 hours a day to complete the safety testing, having no time to double-check the results. Altman wanted to deploy the GPT-4o model before the Google I/O 2024 event, in a bid to upstage the search giant.
OpenAI labeled the GPT-4o model safe enough to deploy and it was released a day before Google I/O. However, later investigations found that GPT-4o surpassed OpenAI’s standards for persuasion. Researchers worked on the fix after the model was released. OpenAI still maintains that GPT-4o is safe enough to deploy.
In all of this, Mira Murati, the then CTO of OpenAI, reportedly pushed back the release of many products including SearchGPT and ChatGPT Advanced Voice mode citing they were not ready for launch.
Sutskever Never Returned to the OpenAI Office
Ilya Sutskever was the one who led the coup and fired Sam Altman last year. In fact, Geoffrey Hinton, Sutskever’s doctoral supervisor who recently won the Nobel Prize said, “I’m particularly proud of the fact that one of my students fired Sam Altman.“
Once Altman was back at the helm, Sutskever never returned to OpenAI’s office. According to the WSJ report, Altman along with Brockman tried to bring Sutskever back, bringing cards and letters from researchers and engineers, requesting him to return.
At one point, Sutskever suggested that he would come back, but Greg Brockman later informed Sutskever that OpenAI was calling off the offer. Ilya Sutskever formally resigned from OpenAI in May 2024 and started his own venture called Safe Superintelligence Inc. in June.
Disappointed by OpenAI’s failure to bring back Sutskever, OpenAI co-founder John Schulman quit the company and joined the rival AI lab, Anthropic. Schulman was also frustrated over internal conflicts at OpenAI and the company’s departure from its original mission.
Sam Altman Under Spotlight
OpenAI, under the leadership of Sam Altman, just raised $6.6 billion in the largest funding round ever, bringing the company valuation to a staggering $157 billion. The investment was primarily backed by Microsoft, Nvidia, Thrive Capital, and other VC investors. Apple backed out of the funding round at the last moment.
The Financial Times reported that during the funding round, OpenAI asked investors to not fund rival AI startups including Anthropic, Elon Musk’s xAI, Perplexity, and Sutskever’s SSI. It seems Altman is truly going out of his way to block the competition.
In contrast, OpenAI’s Charter says “if a value-aligned, safety-conscious project comes close to building AGI before we do, we commit to stop competing with and start assisting this project.“
While Altman is all-out to block the competition and ship products that will attract billions in investment, its founding members, researchers, and top scientists are leaving the company.
It will be intriguing to see how long OpenAI can maintain its technical leadership in the AI industry in the coming months. But for now, we can say that by shifting its focus from research to products, OpenAI has strayed from its original mission.
