HOT TAKES

Current as of March, 2026

Digital Assets

Programmable money is useful.

We see tokenization overtaking nearly every sector of international finance and payments. We believe the US and a few other agile and able jurisdictions will capture much of the value of tokenization—in the US through institutional capacity building, elsewhere through regulatory arbitrage—and create a new global layer that sits above individual nation-state-backed reserve currencies.

Round one of the digital asset innovation era is done. Much of the fundamental infrastructure has been built, and the next phase will see an explosion of value for infrastructure providers and gap-bridging platforms reliant on that infrastructure.

We believe privacy, self-sovereignty, and permissionless innovation must continue to be at the forefront of design for digital assets to truly benefit every person in the world.

AI

We have not yet begun to infer as a species.

There will be demand for 100,000x as much inference as there is now over the coming years. While we will not stop discussing whether or not AGI is here, it will no longer matter. Agentic AI and the next level of autonomy will replace jobs at massive scale, and simultaneously make an immense amount more work for the small percent of the world's population that innovates.

Physical robots will come, at scale, soon. They will transform many sectors of the economy in the US and elsewhere.

We think datacenters in space might actually be real. Regardless, energy costs will rise massively.

Microsoft's aim to serve sovereign inference needs is smart, and we think the market is dumb for ignoring it.

Countries have spent the last few years sorting into haves and have-nots as viewed through the lens of local frontier model teams. Right now, the United States and China are the only key jurisdictions with frontier model capacity. Chinese models fast-follow in part because of significant data exfiltration from US teams. Europe is searching for an answer. India has finally begun to try to build its own frontier lab, but outcomes are not certain.

There is a significant risk to nation-states ignoring this arms race. We believe it may materially impact national economies by 2028.