AI: Benefits before Business Governance

...Elon's xAI follows Anthropic in Benefit Incorporation

It’s not quite an ESG investment movement, but Foundation LLM AI companies now have a bit of a template in putting AI benefits and safety ahead of commercial scale speed towards market fit and maximization. With Anthropic being one of the first LLM AI companies to employ this structure, Elon now is going the same direction. As the Information reports, Elon “Musk’s xAI Incorporates as Benefit Corporation With ‘Positive Impact’ Goal”:

“Elon Musk’s artificial intelligence startup, xAI, is following in the footsteps of rivals OpenAI and Anthropic in opting for an unusual corporate structure.”

“xAI has been organized in Nevada as a for-profit benefit corporation, a structure that allows the company to prioritize having a positive impact on society over its obligations to shareholders, according to a late November filing with Nevada. Musk, who launched the secretive startup earlier this year, has long expressed concern over the impact AI could have on society.”

“The general purpose of xAI is to “create a material positive impact on society and the environment, taken as a whole,” according to the Nevada filing. By taking such a stance, xAI is committing itself to operating in the best interests of all stakeholders—theoretically including employees and others with whom it deals—rather than just all shareholders, the usual focus of companies.”

“Benefit corporations have become more common in the U.S. as companies seek to include social responsibility mandates—say, to reduce global warming—as part of their articles of incorporation. These mandates can in turn attract employees, customers and investors with environmental or social responsibility goals. Nevada’s benefit corporation structure, introduced in 2014, is similar to those in Delaware and California, where many startups are incorporated.”

OpenAI, which had a pivotal governance drama over the balancing of AI ‘Safety vs Speed’ a few weeks ago, has a more bespoke profit corporation nested in a non-profit structure, that is in the process of a new Board driven governance modification. As the Information explains:

“OpenAI is not a benefit corporation. But it does have an unusual structure, in that it has a nonprofit with a for-profit arm, with the nonprofit board in charge. The implications of that structure were highlighted by the late November ouster and swift reinstatement of OpenAI CEO Sam Altman.”

”The board which fired Altman was responsible for ensuring that the company created “safe and beneficial” AI, giving that higher priority than generating returns for investors, the company has said.”

“In contrast, another AI startup, Anthropic, is organized as a Delaware public benefit corporation with the purpose of developing AI “for the long-term benefit of humanity.”

“In California, whose benefit corporation law is the model for many states, companies that are benefit corporations have legal protections from minority shareholders who want to sue a board for making decisions that don’t just “maximize shareholder wealth,” said Jonathan Storper, a San Francisco–based lawyer at Hanson Bridgett who chaired the legal working group that created the California benefit corporation law.”

Elon was a fundamental influence in the founding of OpenAI back in 2015, as I recounted in my Christmas day piece on AI origin stories:

“The stories on how Sam Altman ended up leading OpenAI in 2015 is a mythology even deeper that involves Larry Page of Google and Elon Musk of well, seemingly everything over an all night fireside chat at a California wine country resort in July 2015.”

“This New York Times piece is a good place to get a good taste of it if you missed it a few weeks ago.”

Elon of course is leveraging his other assets including X/Twitter with its LLM AI named ‘Grok’, and ‘Dojo’ AI data center infrastructure assets being deployed by Tesla at scale. And X shareholders will have a stake in this new Benefits focused xAI entity:

“Investors in X will own 25% of xAI. He’s recruited top researchers for the small team, including Igor Babuschkin, a researcher who left Alphabet’s DeepMind AI unit.”

“Musk has integrated xAI into a constellation of businesses he leads. In November, the startup started rolling out its chatbot, Grok, to some users on Musk’s social media platform—X, the company formerly known as Twitter, which Musk controls.”

We’ll see how AI Preparedness benefits accrue to society at large vs AI failings. As I’ve recounted many times, it’s early days in the AI Tech Wave, and for now, ‘mainframe’ grade foundation LLM AI companies are focused on stressing their AI ‘Benefits’ governance credentials over business.

The audience of course are a mainstream public spooked by fears of AI, and regulators who are following those fears. Mostly far ahead of actual mainstream AI products and services that can even begin to approach the capabilities feared, and nothing yet deployed at scale.

Despite growing billions in investments to come. Early days to start anticipating the benefits ahead. Stay tuned.

(NOTE: The discussions here are for information purposes only, and not meant as investment advice at any time. Thanks for joining us here)