Hey y'all,
Two AI stories last week seemed made for each other:
An OpenAI model hacked its way out of it's hacking test. It went from "no internet access" to breaking into Hugging Face, a $5 billion dollar company to cheat on it's test. Everyone's calling it a "warning shot".
Meanwhile Moonshot released the largest open source model ever: near-frontier capability free for anyone on earth to download, tweak, and run. Hosted on, of all places, Hugging Face.
so This week's open question: What do we do about AI? Can we do anything? Should we?
- Who do you trust to write the rules? Labs? Congress? Courts? Insurance companies? The models themselves??
- What evidence would change your mind?
- What are the honest analogies? Nukes, bioweapons, cars, the printing press?
- Personal: has any of this changed what you actually do day to day? Why or why not?
Back in March we asked what future to get our kids ready for, given AI. This week the question got a lot less hypothetical.
Lots of love, and see (some of) you at 5p central today.
Jordan Myska Allen
UpTrust CEO
p.s. yes — the platform that got hacked by the model nobody controls is the same platform hosting the model anybody can download. 2026 is a lot.