AI summary by TwinMind
The Entire OpenAI Chaos Explained
ColdFusion
15m 33s
Initial Firing and Immediate Aftermath
- Sam Altman fired as OpenAI CEO on November 17, 2023
- Fired via Google Meet call while in Las Vegas for Formula 1 race
- Board members present except Greg Brockman (co-founder)
- Accused of being "not consistently candid in his communication with the board"
- No clear reason given, immediately locked out of computer
- Greg Brockman (co-founder) demoted but asked to stay
- Quit immediately in protest
- Tweeted support for Altman
- Mira Morati (CTO) appointed interim CEO
- Only person who knew about firing plan beforehand
- Microsoft CEO Satya Nadella expressed support for Altman
- Microsoft has $13 billion investment, 49% stake in for-profit wing
- No board seats, could only watch from sidelines
Employee Revolt and Chaos
- 747 of 770 employees sent open letter to board
- Ultimatum: "resign or we're all going to Microsoft"
- OpenAI headquarters stood deserted
- Rumors circulated about Microsoft extending offer to Sam and Greg
- Employees plotting Altman's return
- Mira Morati turned against board, joined "team Sam"
Board Changes and Failed Negotiations
- Altman met with board day after firing to discuss return
- Meeting went late into night
- Board did not change position, Sam still out
- Ilya Sutskever (chief engineer who orchestrated firing) made u-turn
- Expressed regret on Monday
- Aligned himself with Sam
- Emmett Shear (former Twitch CEO) appointed new CEO
- Surprised by appointment
- Pledged to bring clarity and initiate independent investigation
- Served only 72 hours as interim CEO
Altman's Reinstatement
- Sam Altman got job back November 22nd (5 days after firing)
- New board formed with Brett Taylor (former Salesforce co-CEO) and Larry Summers (former treasury secretary)
- All old board members left except Adam D'Angelo
- Greg Brockman also returned
- Strong partnership with Microsoft continues
OpenAI's Unique Corporate Structure
- Founded 2015 as nonprofit with mission to benefit humanity
- Celebrity founding team included Elon Musk (left 2018 due to conflict of interest)
- For-profit arm established to fund expensive AI research/development
- Nonprofit faction maintains ultimate power through board control
- Structure left Sam and Microsoft at mercy of safety-conscious board
Reasons for Firing - Side Projects and Conflicts
- Sam working on projects beyond OpenAI board's comfort zone:
- Worldcoin: crypto venture using eyeball scanning, marketed as AI job loss solution/UBI stepping stone
- AI chip venture to reduce Nvidia reliance, seeking tens of billions from Middle East sovereign wealth funds
- Partnership with Johnny Ive (former Apple designer) for AI hardware company, pitched to SoftBank
- Board saw these as distractions from OpenAI's core mission
- Conflict with Ilya Sutskever over superintelligent AI control team formed in July
- October: Altman attempted to reduce Sutskever's role in company
Developer Conference Trigger
- November 6th: OpenAI's first developer conference in San Francisco
- Sam announced customized ChatGPT versions (custom GPTs)
- Task-specific chatbots that might operate independently
- Major red flag for safety concerns
- "Last nail in the coffin" for board's patience
Possible AGI Breakthrough Theory
- Reuters report: staff researchers wrote internal letter to board
- Warning about discovery of advanced AI potentially threatening humanity
- Project called "Q*" acknowledged by OpenAI
- Researchers flagged potential danger but didn't specify exact safety concerns
- Some believe this could be anticipated AGI (artificial general intelligence)
- AGI defined as AI capable of outperforming humans in any economically viable task
- Altman hinted at significant breakthrough at San Francisco summit week before firing
Aftermath and Implications
- Microsoft and Sam Altman appear to be biggest winners
- Altman continues leading with theoretically more supportive board
- Internal drama doesn't bode well for company setting AI industry standards
- Governments and policymakers watching closely
- Elon Musk's response to Sutskever's regret tweet: "if OpenAI is doing something potentially dangerous to humanity the world needs to know"
