It has become clear that Ilya Satskiver, the former chief scientist at OpenAI who led the resignation of CEO Sam Altman, had distrusted Altman since 2017



In February 2024, Elon Musk, who owns Tesla, SpaceX, and X (formerly Twitter), filed a lawsuit against OpenAI, which develops chat AI 'ChatGPT,' claiming that 'OpenAI has become a de facto closed-source subsidiary of Microsoft, the world's largest technology company.' In the lawsuit, email exchanges between Musk, OpenAI CEO Sam Altman, and former OpenAI chief scientist Ilya Satkivr were made public.

gov.uscourts.cand.433688.32.0_1.pdf
(PDF file)

https://storage.courtlistener.com/recap/gov.uscourts.cand.433688/gov.uscourts.cand.433688.32.0_1.pdf

New OpenAI emails reveal a long history of mistrust
https://www.transformernews.ai/p/openai-emails-altman-trust



Internal emails: Elon Musk wanted to keep OpenAI from becoming 'Microsoft's marketing bitch' – GeekWire

https://www.geekwire.com/2024/internal-emails-elon-musk-wanted-to-keep-openai-from-becoming-microsofts-marketing-bitch/

OpenAI Email Archives (from Musk v. Altman) — LessWrong
https://www.lesswrong.com/posts/5jjk4CDnj9tA7ugxr/openai-email-archives-from-musk-v-altman

Musk filed suit against OpenAI in San Francisco Superior Court on February 29, 2024 (local time). In the complaint, Musk criticized that OpenAI was originally founded as a non-profit organization to develop open source AGI (artificial general intelligence), but has since transformed into a de facto closed-source subsidiary of Microsoft. He also pointed out that OpenAI's large-scale language model 'GPT-4' has 'become not open source software, but closed source software primarily to serve Microsoft's proprietary commercial interests,' and asked for an injunction prohibiting Altman and others from using OpenAI or its assets for financial gain or for the benefit of Microsoft or other entities.

Elon Musk sues ChatGPT developer OpenAI, claiming it is a 'de facto subsidiary of Microsoft' - GIGAZINE



In November 2024, Musk filed an amended complaint with the court, which included email correspondence with Altman and others from 2015.

In one of the emails released by the company, Musk wrote that in 2016, OpenAI met with Microsoft to discuss a partnership, and that 'the terms proposed by Microsoft made us sick. We should have paid at least $50 million to avoid being perceived as a de facto subsidiary of Microsoft.'



In 2017, Greg Brockman, co-founder of OpenAI with Satkiver, told Altman, 'I don't trust Altman's judgment,' and 'Is your true goal the development of AGI? How does the development of AGI tie into your political goals?' At the time, OpenAI was in discussion about who would be in charge of running it as its representative and whether it would convert from a non-profit organization to a for-profit organization. While Musk wanted OpenAI to be a non-profit organization, Altman expressed his intention to convert to a for-profit organization.



'Although you have stated that you do not want to control the final AGI, we are concerned that you will maintain absolute control over the company as OpenAI progresses toward developing AGI,' Satskivar and others said in a statement to Musk.

In response to this email, Musk replied, 'Enough is enough. Either you go independent and start another business, or you continue OpenAI as a non-profit. I'm not going to fund OpenAI until you can firmly commit to staying with it. End of discussion.'

Below is Musk's response to this email being made public.




Later, Shivon Gillis, who was close to Musk, told Musk, 'Mr. Ilya and Mr. Brockman have told me that they want to continue OpenAI's non-profit structure, and after talking with them, Mr. Altman has decided to take 10 days off to reconsider. Leaving OpenAI will allow me to see how much I trust them and how much I want to work with them.'

Satkiver subsequently reportedly led Altman's departure and exit from the company in November 2023 as a result of conflicts with Altman.

Is OpenAI's former CEO Sam Altman's resignation due to a dispute with chief scientist Ilya Satskiver over 'AI safety'?

in Note, Posted by log1r_ut