Visions of Freedom

ALTMAN FIRED: The Prematurely Reported Death of Mr. Altman – 11/17/23

Altman FIRED OpenAI Guest Badge selfie tweet

BREAKING NEWS // TUES // NOV 21, 2023 // 10PM PST
SAM ALTMAN RETURNS TO OPENAI IN TRIUMPH. 
TONER & MCCAULEY EJECTED FROM BOARD.
NEW BOARD ESTABLISHED. NADELLA GOES INTO THERAPY.
SUTSKEVER GROVELS & PRAYS. BACK TO WORK.

Eric Schmidt sums it up:

“It’s pretty simple:
The board fires Sam.
Sam fires the board.“

<TL/DR>

BREAKING NEWS // FRIDAY // NOV 17, 2023
SAM ALTMAN FIRED FROM OPENAI. BROCKMAN QUITS.
SUTSKEVER DENIES COUP. BRAIN DRAIN BEGINS.

Sam Altman Fired from OpenAI / key points:

Friday 11/17:

  • Sam Altman fired without stated reason as CEO of OpenAI
  • co-founder Greg Brockman (President & Chairman) quits in solidarity 3 hours later
  • Microsoft stock falls 2 full points (-2%) on the news. That’s close to –$50B in market cap. Oops.
  • Mira Murati (CTO) named interim CEO
  • stated corporatespeak rationale:
    Altman was apparently “not consistently candid in his communications” with the board and therefore “the board no longer has confidence in his ability to continue leading OpenAI,” — these were the only (non) reasons given.
  • read OpenAI’s official blog post regarding the ouster.
  • by midnight Friday, 3 senior researchers at OpenAI had also tendered their resignations

Saturday 11/18:

  • investors threatened to sue,
  • Microsoft (allegedly) threatened to withhold Azure Cloud Compute credits, and
  • massive pressure was placed on the Board to reverse their decision and re-instate Altman

Sunday 11/19:

  • negotiations to bring Altman back fall apart
  • the board names a new interim CEO: Twitch co-founder Emmet Shears
  • in a stunning midnight grandmaster chess move, MSFT CEO Satya Nadella announces:
    • “Sam & Greg are joining Microsoft as employees
      to form a new fully funded AI research division.”

Monday 11/20:

  • MUTINY!!!!
  • fully 720+ of OpenAI’s 770 employees (>90%) co-sign a letter demanding that:
    • the board resign,
    • that Altman be re-instated as CEO, or else:
    • they would “imminently” quit and agree to work for Microsoft at the new AI R&D division.
  • Two key results of this:
    • a) MSFT essentially acquires the key assets (humans) of a high-flying $80B company for… a net $11B
    • b) OpenAI is left with a 4 person board,
      • less than 20 employees, and…
      • $10 billion cash in the bank (not bad), and…
      • a slew of lawsuits that will take 4 years to work their way through the courts (bad)

Tuesday 11/21:

  • Altman has not formally accepted MSFT job offer;
    • appears to be using it as a negotiating ploy / counterweight as he…
  • continues to actively negotiate with OpenAI board
    • regarding his & Brockman’s return
    • new governance structures
    • and the resignation / replacement of the current board
  • 10PM : IT’S OVER. ALTMAN IS BACK.

Basic Analysis: a Coup d’état

  • By all accounts, this was a uni-lateral decision by the board (not including Brockman, the chairman, who was excluded). Altman fired, Brockman, OpenAI’s roughly 600 employees, and even both the CEO and CTO of Microsoft, who is on the line with a $13 billion investment and deep product licensing ties, were all apparently taken by total surprise.
  • MSFT (Micrsoft stock) dipped close to 2 full points on the news. That’s minus $50 billion on their current market cap. Oops.
  • All signs point to a coup d’etat orchestrated by Ilya Sutskever, and supported by the (3) external board members.
  • Sutskever is a co-founder and Chief Scientist of OpenAI. He is a prodigy of Geoffrey Hinton, the “Godfather of AI.” Sutskever’s research on scaling (bigger datasets, trillion parameter models, longer training runs) os the primary driver behind the past few years of insane innovation in the AI industry.
  • The core schism between Sutskever and Altman appears to be what is known in AI circles as “decel vs. accel”: put simply, Altman was constantly pushing to release more capable / intelligent models, faster, to more consumers. Sutskever used to be on this “accel” bandwagon, but 3 months ago decided that AI was posing existential risks to humanity, and devoted 100% of his energies to AI safety research.
  • OpenAI has an utterly unique corporate structure (details here), where a 501(c)3 non-profit fully controls a “capped profit” subsidiary, which can (and does, spectacularly) receive traditional VC investment.
  • So you have a non-profit, with a skeletal board comprised of mostly academics, controlling a for-profit company building the world’s most advanced AI, using more than $13 billion ($13,000 million) raised from Microsoft (and Elon Musk).

Who is the Board of OpenAI?

The board, as of November 2023, consisted of just 6 individuals, 3 of which were insiders (employees). It is of note that for a high-flying Silicon Valley tech startup with an astronomical $80 billion valuation, this board is considered both lightweight — low-Q, no major Valley VC or keiritsu founders (other than d’Angelo) — and skeletal (too much power in the hands of way too low a number of people — just six):

  •  Ilya Sutskever @ilyasut
    • Chief Scientist @OpenAI
    • prodigy of Geoffrey Hinton, the “Godfather of AI”
    • co-author of the seminal AlexNet paper, 2012
    • Sutskever announced in August 2023 that henceforth, he would be dedicating 100% of his energies to AI Safety. In a nod to this, OpenAI announced that a full 20% of their supercomputer / datacenter capacity would be devoted to safety research & guardrails.
    • Russian-born Israeli-Canadian
    • wikipedia
  • Adam d’Angelo @adamdangelo
    • career: first CTO of Facebook (departed 2008), founder of Quora
    • education: Cal-Tech
    • joined OpenAI board in April 2018
    • wikipedia
  • Helen Toner @hlntnr
    • CSET
    • education: Georgetown
    • a well respected academic,
    • CSET is a key institute that advises PotUS on AI / cybersecurity policy.
  • Tasha McCauley @tashamccauley
    • a total wildcard
    • education: MBA from USC Marshall School of Business.
    • who the hell is this woman, 1/4 of the voting power controlling an $80B juggernaut?!?!?
    • mainstream media enigmatically labels her a “tech entrepreneur”… when they even bother to mention her at all.
    • CEO at GeoSim, co-founder of Fellow Robots — neither of these companies have valid web presences.
    • curiously, married to Joseph Gordon-Levitt (Inception)
  • [x] Greg Brockman quit @gdb
    • co-founder, (ex) President, Chairman of the Board @OpenAI
    • education: Brockman briefly attended Harvard, later dropped out of MIT in 2010
    • career: employee #4 at Stripe, its first CTO
    • co-founded OpenAI in 2015 alongside Sam Altman & Elon Musk
    • wikipedia
  • [x] Sam Altman fired @sama
    • co-founder, (ex) CEO @OpenAI
    • former head of Silicon Valley’s legendary Y-combinator
    • the global poster child of AI
    • perhaps the most powerful founder in Silicon Valley, c. 2023

Who are the Alumni of OpenAI’s Board of Directors?

It didn’t always used to be this way (a skeletal board of 4 obscure personalities). In fact, OpenAI’s board has hosted some of the most powerful individuals in Silicon Valley, including Elon Musk and Reid Hoffman. Here’s a breakdown of historical board members, their dates of tenure, and their reason for departure:

Tasha McCauley Adam d'Angelo Helen Toner OpenAI Board of Directors

Elon Musk, Holden Karnofsky, Reid Hoffman, Shivon Zilis, Will Hurd, Greg Brockman, Sam Altman

Active theories I am cultivating about why Altman was canned:

(this section was written off the cuff Friday afternoon,
mere minutes after the news hit of Altman’s termination.
It now appears to — possibly — be all about the
religious AI divide of Safety vs. Speed to market,
a.k.a. Decels vs. Techno-Optimists)

1) (most probable) Big Money got fed up with the dreamy-eyed children. They came in and said:

“playtime’s over, kids. all that utopian philosophy you espoused about ‘capped profit’ and UBI was nice, but this is world-changing. this is real business. and business is war. buh bye, thanks for all the fish, we’re bringing in the seasoned generals & combat troops now.”

altman-fired-openai

2) (long shot) Altman and Brockman withheld knowledge about an actual AGI / ASI that was ready to take over the world — or at least Wall Street — that OpenAI had developed. Certain board members, lacking proper insider information, were offended. They took action.

3) (The most mundane)  Sama made some off-hand comment in public (he was never afraid to speak his mind nor change positions), that just “crossed the line” too far of someone’s offense metrics.

PS – kicked Brockman’s Chairman title outta here, same as @sama.
Something stinks in Denmark…

 

Altman is being very polite about it… for now:

altman fired. final tweet

Surviving Board members of OpenAI after Altman fired are:

  • Ilya Sutskever, OpenAI’s chief scientist
  • Adam D’Angelo, Quora CEO & Facebook co-founder
  • Tasha McCauley, tech entrepreneur
  • Helen Toner, Dir. of Strategy at Georgetown University’s CSET.

Best Twitter (X) Perspectives on the Coup:

Who broke the story?

The original story of Sam Altman Fired,
as reported by WSJ:

altman fired

prompt: a political cartoon, style of new yorker. a hooded executioner, with the nametag “Microsoft” holding a scythe in one hand, and has a speech bubble above that reads “Utopias Over, Kid.”. there is a guillotine with a prisoner kneeling with his head on the chopping block, wearing a nametag that says “altman.”.

engine: DALL-E3, OpenAI

AcroYogi

, ,