Half a trillion $, Grokipedia, & the ultimate heist

TLDR;

Welcome back! This week, we have a lot of news from OpenAI, NVIDIA, Qualcomm, and Alibaba, the latter of which may not be welcome news for investors.

Additionally, we’ll take a look at Elon Musk’s new tool in its crusade against the left, Anthropic’s new Excel add-in, and the next potential AI IPO.

Enjoy!

OPENAI

A lot to talk about with OpenAI

OpenAI has completed a major restructuring aimed at paving the way for a future IPO that could raise tens of billions of dollars.

The confirmation of the restructuring still hinges on the Elon lawsuit, which goes to trial in March.

The reorganization converts all prior investments into regular equity while keeping control under a nonprofit board, the OpenAI Foundation. Microsoft now holds a 27% stake valued at about $135 billion, while the foundation owns 26%, valued at about $130 billion. Early investors saw dilution, but the changes remove previous caps on financial returns (previously, investors could only hope for a 100x return; now it’s uncapped).

The foundation retains ultimate control of OpenAI’s governance, ensuring the company adheres to its mission of benefiting humanity (sure). Its board can overrule the corporate board on safety matters, and OpenAI pledged to keep its safety and security committee independent.

Regulators in Delaware and California approved the restructuring after confirming these commitments. The foundation could receive further shares if OpenAI ever exceeds a $5 trillion valuation, with plans to use proceeds for research on AI-related social risks.

TheWhiteBox’s takeaway:

The highlight for me, besides the fact that this is, making no distinction, a heist on US taxpayers (more on this in a second), OpenAI CEO Sam Altman said a public listing is likely due to the company’s massive capital needs.

I called it a few weeks ago; there’s really no other option than to IPO if you’re looking to spend $1.4 trillion, as Sam himself shared.

But yeah, this is daylight robbery. The company was incorporated as a non-profit, paying no taxes for years. Now, once it’s printing money, it no longer wants to be a non-profit and, in turn, become a for-profit.

The question here is straightforward: what prevents startups from doing the exact same thing going forward? Constituting as non-profits, grow without tax burdens, and then converting?

And more importantly, how is a company constituted to develop AGI for the “benefit of all humanity”, be a for-profit? (This applies to all other AI Labs by the way).

We always knew the rules of the game were never meant to be fair, but I don’t think we realize the questions (and headaches) this will create.

On a final note, they shared their research roadmap goals:

Commitment to a lot more compute, with ~30 GW of power already in the roadmap

Expecting to have the equivalent of a junior AI researcher-level AI by September 2026

Having a fully-fledged AI researcher, an AI capable of pushing the frontier of AI forward, by March 2028 (oddly specific by the way).

HARDWARE

Qualcomm Presents AI…

Enjoyed the introduction? Subscribe or sign in to continue with this edition.