OpenAI's Sam Altman out—maybe
Newsflash: Over the weekend, and with a lot of pressure coming from Microsoft, investors, and company employees, Sam Altman was reportedly in negotiations to return to OpenAI as CEO. The Information obtained a memo to OpenAI staff from Jason Kwon, the company’s chief strategy officer, saying the company was ‘optimistic’ about reinstating Altman, cofounder Greg Brockman, and other senior staff members who resigned in protest. The Information further reported that Interim CEO Mira Murati earlier told staff that she and the leadership team invited Altman over to the company.
However, the cards fall here; many of our observations below, such as Microsoft’s overly-dominant role and the need to restructure the company and disconnect it from the non-profit, are becoming even more evident.
OpenAI RIP
No one should be ‘shocked’ by the upheaval at OpenAI and the ouster of its CEO, Sam Altman. We’ve been expressing our doubts about OpenAI prospects all along. When chatGPT was launched just short of a year ago, OpenAI was the rage, and its AI Kingship seemed all but enshrined. In Silicon Valley, we learned the hard way that ‘the first mover advantage’ is largely a myth.
The most dramatic example is Google’s journey to search engine dominance (91.6 percent market share) over the backs of the eleven VC-backed search engines that came before them.
The Search Engine Wars during the Roaring ’90s
Google outpaced rivals by introducing the search rating algorithm (PageRank) based on search relevancy and the “pay-per-click” model ‘borrowed’ from Bill Gross’s GoTo.com—later called Overture Services.
Excite: Founded in 1993, —Total VC $81M —Top VC investor(s): Kleiner Perkins, Polaris & IVP. Excite had its IPO in 1996, and later merged with @Home Network in 1999, and ultimately popped with the Internet bubble.
Yahoo! Search: Started in 1994—Total VC $ —Top VC investor(s): Sequoia Capital. Yahoo had its IPO in 1996, and Verizon acquired the company in 2017.
Infoseek: Started in 1994—Total VC $24.7M —Top VC investor(s): Battery Ventures and Menlo Ventures. The Walt Disney Company later acquired the company.
Lycos: Spun out of Carnegie Mellon University in 1994—Total VC $36M —Top VC investor(s): CMGI, Highland Capital & @Ventures. Lycos had its IPO in 1996, and Terra Networks acquired the company in 2000.
WebCrawler: Started in 1994—Total VC $ —Top VC investor(s): The company AOL later acquired the company.
AltaVista: Founded in 1995—Top VC investor(s): Digital Equipment Corporation (DEC).
Ask Jeeves: Started in 1996—Total VC $61M —Top VC investor(s): Highland Capital Partners, IVP, Redpoint Ventures, VantagePoint, & Advent.
Inktomi: Started in 1996—Total VC $ —Top VC investor(s): Oak Investments, Integral, & Intel. Inktomi had its IPO in 1998 and sold to Yahoo in 2003.
Overture (GoTo.com): Founded in 1997—Total VC$: $132M—Top VC investor(s): Draper Fisher, IdeaLab (Bill Gross).
Google: Started in 1998—Total VC $ —Top VC investor(s): Andreas Bechtolsheim, Sequoia Capital, Kleiner Perkins, Shaquille O'Neal, Tiger Woods & Arnold Schwarzenegger.
In other words, chatGPT seemed really cool last November, but they face a growing number of competitors, including Elon’s new Grok, which will be fueled by hundreds of millions of X members, and Google Bard, with its search query monopoly advantage. History would dictate that OpenAI's fate would be like that of the early search pioneers, with too many arrows in its back to remain standing.
Ineffective Altruism
In December 2015, Elon Musk, Greg Brockman, Sam Altman, Reid Hoffman, Jessica Livingston, Peter Thiel, Amazon Web Services (AWS), Infosys, and YC Research formed the non-profit OpenAI with the vision of ‘ensuring that artificial general intelligence (AGI) benefits all of humanity.’ Combined, they pledged over $1 billion to the AI non-profit venture, of which the company ultimately only collected $130 million.
In 2018, Elon resigned from his board seat, citing 'a potential future conflict of interest' with his role as CEO of Tesla due to Tesla’s AI development for self-driving cars. Sam Altman claims that Elon believed OpenAI had fallen behind other players like Google, and Elon proposed to take over OpenAI himself instead, which the board rejected.
To compete for AI talent against deep-pocketed players like Google Brain, DeepMind, or Facebook, OpenAI spun out a for-profit entity in 2019 to be able to attract investment from venture funds and compete for talent by granting them stakes in the company. The one atypical twist in the new structure is all VC returns were ‘capped’ at 100X.
We are neither lawyers nor have any particular inside knowledge of the structural affairs of OpenAI. Still, as close observers, we had the following gut concerns about the company.
Convoluted corporate governance structure. OpenAI’s unusual structure gave the non-profit’s six-member board the power to govern the activities of the capped-profit entity, including firing the CEO—which they did. Also, under this structure, Sam nor anyone on the board directly owned any shares in OpenAI.
‘For-profit’ companies that have investors and employees looking for an ROI deserve industry-seasoned board members whose primary fiduciary responsibility is maximizing stockholder value while maintaining proper governance practices.
“The reason OpenAI exists at all is because I used to be really close friends with Larry Page and stay at his house in Palo Alto. We would talk late in the night about AI safety, and my impression was that Larry wasn’t taking AI safety seriously enough. He really seemed just focused on achieving digital superintelligence—essentially a Digital God, if you will, and as soon as possible. This is not good. So I thought, ‘What is the furthest thing from Google?’ which would be a fully open nonprofit. So the ‘open’ in OpenAI stands for open source and transparency so people know what is going on. I’m normally in favor of for-profit companies, but the idea was not to be a profit-maximizing demon from hell that never stops. So that is why OpenAI was founded. Very, unfortunately, they decided to become a for-profit company.”
—Elon Musk’s interview with Tucker Carlson last April.
There also appears to be an ideological battle going on between the company and the non-profit board overseeing it. From what we can tell from public remarks by Sam, cofounder Greg Brockman, and even new temp CEO Mira Murati, the company is highly sensitized to the potential perils of AI. Ms. Murati, in particular, has articulated very practical approaches to implementing internal safeguards and creating industry-accepted standards. On the other hand, at least two board members (who voted for Sam’s ouster) have ties to the Effective Altruism movement, a utilitarian-inspired group that is known to stoke fears that AI development will create a Terminator-type scenario where we all become slaves to the robots. As radical as it sounds, even Elon harbors these fears, and these were, in part, why he left OpenAI—because he was looking for faster action.
We are not students of Effective Altruism, so we wouldn’t know if its tenets have anything to do with chatGPT’s preachy epilogues OpenAI tacks on to searches that reek of Web2-style manipulation and censorship. As Elon observed, ‘Humor is clearly banned at OpenAI, just like the many other subjects it censors.’ Plus, its responses are poorly written, often repetitive, and filled with incorrect facts. What we are looking at here, folks, is a BMVP (barely minimum viable product), with competitors coming after OpenAI like heat-guided missiles.
Sam Alton seems like a nice enough fellow, but he always came across to us as more of a Y Combinator (his previous gig) wheeler-dealer type than a maniacally focused change-the-world entrepreneur. Leading a pioneering AI company that has the world’s attention is at least three full-time jobs. What’s up with Sam’s other take-over-the-world project—Worldcoin? And is Sam the guy we think is going to create the first AI phone?’ Sam should go back into retirement and be a VC—and sit on OpenAI’s board, where he can still help them network.
We also have had concerns over OpenAI being overly tied to outside corporate interests, specifically Microsoft, Amazon, and Softbank. OpenAI is very expensive to operate and can’t afford to be led around by a tail of large commercial partners who distract the company from its top priorities.
The bottom line here is that OpenAI is so riddled with structural, personal, and ideological differences that it breaks way too many of the tried-and-true governance principles of a successful Silicon Valley startup. Read Bill Gurley’s Venture Capital Red Flag Checklist for a deeper understanding of these principles.
OpenAI’s competitors—Here they come, walking down the street
Cohere: Canada-based started in 1998—Total VC $440M—Top VC investor(s): Coatue, Intel, Lightspeed, & Samsung.
Hugging Face: Brooklyn-based founded in 2016—Total VC $400M —Top VC investor(s): Sequoia Capital, Alphabet, Amazon.com, Betaworks, Coatue, IBM Ventures, Intel, NVIDIA, & Salesforce Ventures.
Stability AI: London-based founded in 2019—Total VC $1.0B —Top VC investor(s): Betaworks, Coatue, Lightspeed, & Samsung NEXT Ventures.
Jasper: California-based founded in 2018—Total VC $140M —Top VC investor(s): Bessemer, Coatue, Foundation Capital, Insight Partners & IVP.
Databricks: San Francisco-based founded in 2015—Total VC $4.1B —Top VC investor(s): Andreessen Horowitz, NEA, Microsoft, & DCVC.
DataRobot: Boston-based, founded in 2012. —Total VC $1.0B —Top VC investor(s): Sutter Hill, Accel, BlackRock, Citi Ventures, Glynn, Insight, Intel,, Morgan Stanley, & NEA.
Inflection AI: Palo Alto-based founded in 2022—Total VC $1.5B—Top VC investor(s): Bill Gates, Eric Schmidt, Greylock, Microsoft, Nvidia, Polaris, & Reid Hoffman.
A Hail Mary scenario is the AI-savvy temp-CEO Ms. Murati shows the cajones to stand up and say she is ready to ride this pony to victory on the following conditions (and whatever else she needs to get the job done):
Restructure OpenAI as an independently controlled for-profit entity with no investor cap.
Recruit an experienced board of directors that should include Ms. Murati, Sam, Greg Brockman (OpenAI’s cofounder and president who quit on Friday in solidarity with Sam), a Microsoft representative, a VC investor representative, and at least six other industry-savvy, independent directors who take the job of corporate governance very seriously.
Create industry-standard employee stock option plans and liquidity strategies in order to successfully recruit the best and the brightest in the AI industry.
Each OpenAI executive must sign a pledge that making the company successful is their only professional focus and will be as long as they get an OpenAI paycheck.
Articulate a united and realistic view of the potential risks of AI and how OpenAI intends to straddle these risks while still achieving maximum growth to secure its place in the market. It’s better if a transparent and accountable US company like OpenAI takes the AI lead than the CCP or some otherwise unknown nefarious operation.
There is no place at OpenAI for ‘doomers.’ It appears the coup instigated on the non-profit side was doomer-led—all the more reason to restructure the company and board.
We don’t know Ms. Murati, but we are pulling for her. Otherwise, Microsoft will buy OpenAI and swallow it into the bowels of its bureaucracy, never to be heard from again.
May we all take comfort in knowing that no AI app can or will ever replicate Lady Wisdom. Good luck, Ms. Murati—you are clearly wise enough to do this.
The Book of Wisdom 7:22–8:1
And now I understand everything, hidden or visible, for Wisdom, the designer of all things, has instructed me.
For within her is a spirit intelligent, holy, unique, manifold, subtle, mobile, incisive, unsullied, lucid, invulnerable, benevolent, shrewd, irresistible, beneficent, friendly to human beings, steadfast, dependable, unperturbed, almighty, all-surveying, penetrating all intelligent, pure and most subtle spirits.
For Wisdom is quicker to move than any motion; she is so pure, she pervades and permeates all things. She is a breath of the power of God, pure emanation of the glory of the Almighty; so nothing impure can find its way into her. For she is a reflection of the eternal light, untarnished mirror of God's active power, and image of his goodness.
Although she is alone, she can do everything; herself unchanging, she renews the world, and, generation after generation, passing into holy souls, she makes them into God's friends and prophets; for God loves only those who dwell with Wisdom.
She is indeed more splendid than the sun, she outshines all the constellations; compared with light, she takes first place, for light must yield to night, but against Wisdom evil cannot prevail.
Indeed, she reaches from end to end mightily, and governs all things well.
The Cryptonite Founder's Pitch Program
Founder’s Pitch Program If you are a hustling innovation entrepreneur ready to pitch your story and boost your funding and branding efforts, it might be time for you to do a Cryptonite Founder’s Pitch. Our job is to bring your pitch alive and put it in front of the investors and influencers who matter. The f…
Cryptonite 300 top Wed3 competition
 | A guest post by
|
Who is Sam Alton?
What's ChatGBT?