- Ilya Sutskever, the cofounder and chief scientist of OpenAI, has decided to leave the company.
- He announced his departure on X and said that he is looking forward to a new project.
- Sam Altman has expressed his sadness over Ilya’s sudden departure and thanked him for everything he has done for the company.
On Tuesday (May 14), OpenAI co-founder and chief scientist, Ilya Sutskever said he’s leaving the company after being a key of its wheel for 10 years. He will be replaced by Jakub Pachocki, the company’s director of research.
“I am excited for what comes next — a project that is very personally meaningful to me about which I will share details in due time.” – Ilya Sutskever on X
What’s interesting is that Altman wasn’t aware of this decision until recently. When he was asked about Sutskever’s status on a Zoom call, he said there was no such announcement. He further said that he loves working with Ilya and hopes to keep with him for the rest of their careers.
However, now that he knows, he shared deep regrets about Ilya’s departure and praised his brilliance, warmth, and compassion. Altman said that he will forever remain grateful for his service and will continue to work on the mission they started together.
This is where it gets even more interesting: just hours after Sutskever announced his departure, Jan Leike, co-leading superalignment of GPT-4o also announced that he is leaving the company with a simple “I resigned” post on X. Altman hasn’t yet addressed Leike’s post.
This isn’t the first time a top-level employee has left OpenAI.
- Just a few weeks ago, Diane Yoon (vice president of people) and Chris Clark (head of nonprofit and strategic initiatives), too, left.
- Researchers Leopold Aschenbrenner and Pavel Izmailov who worked on the superalignment team left last month.
- Two other employees working on safety and governance, including Daniel Kokotajlo, have also resigned.
Naturally, so many resignations, one after the other, have sparked a lot of speculation. While the reason for others leaving is unknown, Kokotajlo stated that he doesn’t trust the way the company is handling AI advancement.
At a time when AGI (artificial general intelligence) is evolving, Kokotajlo cannot trust OpenAI to behave responsibly, which is why he’s leaving.
During the November 2023 fiasco, too, when Altman was fired and then hired back a few days later, there was a lot of talk about Sam’s secret Project Q* – an AGI project to develop systems smarter than human capabilities. The OpenAI board was allegedly skeptical about the immense power and potential misuse of the system, which resulted in Altman being removed.
Funnily enough, at this point, everyone except Sam is leaving OpenAI. This only hints that Project Q* might still be equally (or even more) dangerous than it was in November last year. Where’s the EU when you need it, eh?
Sutskever’s Role in the Firing of Sam Altman
Sutskever had a huge role in the firing, as well as the rehiring, of CEO Sam Altman last November. Almost everyone in the tech industry knows of this scandal. About 6 months ago, Sutskever hit the headlines after he voted out Altman as the company’s CEO for his lack of transparency and for pushing technology “too far, too fast.”
However, it was Sutkskever himself who supposedly had a ‘change of heart’ and helped Altman come back. He signed an employee letter, calling for the entire board to resign and for Altman to be reinstated—and so he was, just five days later. Altman rejoined the company, and board members Tasha McCauley, Helen Toner, and Sutskever left the board.
Sutskever also issued an apology for his actions then and said that his intention was never to harm the company. He only wanted to ensure that a powerful technology like AI is handled with care and security.
As Altman rejoined the board, three new members were also announced, including:
- Dr. Nicole Seligman, former EVP and global general counsel of Sony and president of Sony Entertainment
- Sue Desmond-Hellmann, ex-CEO of the Bill and Melinda Gates Foundation
- Fidji Simo, Instacart CEO
ChatGPT-4o
Sutskever’s resignation announcement comes just a day after the company released GPT-40 – the latest version of its flagship product.
It comes with an improved human touch that makes communication seamless. For example, you can have back-and-forth conversations with the tool without waiting for it to finish a statement. You can also share photos with it and have conversations around them. Other features like GPT memory, custom GPT, and web searches were also added.
GPT-4o will be available to all Plus and Team users, as well as to free users—although with a limit on the number of conversations they can have. OpenAI also plans to roll it out for Enterprise users soon.
Our Editorial Process