The leverage held by talent

The drama surrounding Sam Altman’s firing and re-hiring has dominated this week’s news – and for good reason. It’s an incredible story on so many levels, and the fact it’s spread beyond the confines of the tech press is proof of the wider world’s ongoing fascination with everything AI. 

Altman has been reinstated as OpenAI CEO, with the board members who voted him out ousted and replaced. It’s been a whirlwind week, the stuff of a Netflix miniseries (it’s inevitable…), but the OpenAI story has been unique from the beginning. 

Not your average start up

Originally set up in 2015 as a non-profit charity, OpenAI’s initial intention was to research the benefits and dangers of AI. This blog post from 2015 states the company’s aims, including: 

“Our goal is to advance digital intelligence in the way that is most likely to benefit humanity as a whole, unconstrained by a need to generate financial return.”

A board was put in place, including Altman and Elon Musk, and its mandate was to pursue AI “that is safe and benefits all of humanity,” with the long-term goal of building artificial general intelligence. The board’s focus was not to look after investor interests, as at most startups. 

Heavyweights, such as Musk and Peter Thiel, provided funding and brought in the finest AI minds. But in the early years, the fringe tech work it was doing flew, largely, under the radar. 

Curiously capped for profit

As it rolled out products such as OpenAI Gym and iterations of its language model GPT, OpenAI needed cash, primarily to pay for computing power. In 2019, Microsoft invested $1bn and the organisation switched from a not-for-profit to a "capped for-profit” (capped at 100x investment). Employees received equity and an unconventional layer of structuring was added, with an LLP placed between the non-profit and the capped-for-profit subsidiary. 

The board, including founders, Ilya Sutskever and Greg Brockman, and tech glitterati, like Quora founder, Adam D’Angelo, kept ownership of the non-profit division. Microsoft did not get a seat. No investor did. In fact, the capped-for-profit subsidiary, which distributed equity, had little say over OpenAI’s direction.

The board had the power, including hiring and firing members. Hence, OpenAI’s future was in the hands of a diverse group, each with a different definition of how AI “benefits all of humanity”. 

Tinkering on the fringes 

 

OpenAI has always been different. It tested the boundaries of computing. It didn’t care for the bottom line. It released cool things, like language and image models. For years, it worked like a research institute, tinkering away in a dark corner on technology that few truly understood.  

The world has known about AI’s potential for decades, having been deeply studied in academic circles as well as explored in pop culture. In 2015, OpenAI hailed but also warned of AI’s future: 

“…It’s hard to fathom how much human-level AI could benefit society, and it’s equally hard to imagine how much it could damage society if built or used incorrectly.”

But in 2015, the world still, largely, shrugged. We’d seen Terminator and read Economist articles about AI’s potential to automate – but even by the early 2020s, we’d yet to see much real-world proof. Hence, OpenAI was still not making mainstream headlines. 

An explosion of interest 

In late 2022, this changed. OpenAI released GPT-3 wrapped in a chatbot, called it Chat-GPT, and the rest is history. Microsoft invested a further $10bn, seeing Altman and Open AI as its ticket to ride with the Apples and Teslas for decades. The Big Tech AI arms race was on.

Ever since, OpenAI has been the hottest thing in tech, gaining a monster valuation, forecasts that it would be the most valuable company on earth, and everything else that huge press hype and public interest leads to. The sky was the limit – until the chaos and drama of last Friday. 

A falling out between board members, rumoured to have been about Altman’s views on AI safety and the alignment with Microsoft, led to a majority voting for his removal. Microsoft was blindsided, as were the OpenAI staff, with 750 of 770 threatening to resign if Altman didn’t return. The board’s rashness, and the strange corporate structure, had caused chaos. 

Over the past week, it’s been a war of Tweets, staff counteroffers, negotiations and, eventually, the return of the king. Altman was reinstated and a new board replaced those who plunged the dagger in Altman’s back. D’Angelo remains. Greg Brockman, co-founder and president, who quit after Altman was fired, returned, too. Netflix’s scriptwriters will have a field day…  

The power of leverage 

 

For me, amongst the storylines and characters, the episode is a fascinating study of corporate governance and the leverage held by talent. The original board thought it answered to no one – but of that board, only Altman and Brockman now remain. Intriguingly, the structure remains, with the board, not investors, controlling the company’s future. Or so it says on paper. 

And that’s the crucial point – just because a board has control on paper, it doesn’t mean they won’t have to deal with the unintended consequences of rash decision-making. Whilst they didn’t have a seat or a vote, Microsoft inflicted immense pressure when they learned that their guy, Altman, had been removed. Money talks – and a lot of it talks very loudly. 

More importantly, whether company or charity, talent holds the power, not the board. When the OpenAI workforce threatened to quit (buoyed by job offers from Microsoft and rivals, like Salesforce), the board knew the game was up. Even a company as lofty and ideas-driven as OpenAI was nothing without the engineers who swipe in every day. No company is. 

Ultimately, talent decided the outcome, and this tells us a lot about what makes successful organisations and how it doesn’t really matter what’s written in a contract or a charter. Founders, investors, board members and employees must always be wise to the fact that actions have unintended consequences and that leverage is an incredibly powerful force, especially when it is talent pulling on the lever.

Previous
Previous

A whole lot to learn from Munger

Next
Next

Does London need a stock exchange?