Oh hello there, it's been a while. Once again when the world is chaos I end up not writing anything because I'm so busy making notes of things I want to write about and the backlog keeps getting longer because there's a new thing every 5 minutes and I can't keep up. I've noted this in the past and often tried to find a way around it but at this point I think I just need to give up. Similar to the old email-bankruptcy thing, I just need to accept that I'll never be the most up to date commenta...

An upgrade that enhances pessimistic proofs with a new execution proof mode, enabling chain-agnostic interoperability.

I'm sorry for the recent email issues. I'm working through some bugs with my newsletter platform. Composability is great, but it's not a substitute for reliability. If a Web3 platform can't handle basic features like editing, creators will lose trust. Getting Web1 and Web2 basics right is table stakes. Thanks for your patience while I sort it out.

Came across a tool called Synthetic Users that simulates interviews using AI. It reminded me of a project I ran in 2023 using LLMs to generate proto-personas for my collectors. The tech is fast and can be useful in narrow cases, but it is not a replacement for real people. If you want real insight, you still need to watch real users. This post explores where synthetic interviews help and where they quietly fall apart.

The scientific method is arguably one of the greatest engines of progress humanity has ever devised. It gave us medicine, satellites, quantum mechanics, and the internet. Yet, the infrastructure for science continues to feel increasingly out of step with the internet native, network-driven world we live in. The way science is funded, published, peer-reviewed, and distributed often reflects an outdated model, one built for a pre-digital era where access was scarce and authority was centralized...

MIT’s new study says your brain does not work better after using ChatGPT. But the setup all but guaranteed that result. This post breaks down what the study actually tested, why the conclusions are misleading, and what a more honest look at real-world AI use should include. If you care about how people actually think with these tools, not just what happens when they check out completely, this is worth a read.

MIT’s new study says your brain does not work better after using ChatGPT. But the setup all but guaranteed that result. This post breaks down what the study actually tested, why the conclusions are misleading, and what a more honest look at real-world AI use should include. If you care about how people actually think with these tools, not just what happens when they check out completely, this is worth a read.

The intro post of a five-part series on why and how inviting in student voices early has shaped the way I build, test, and think about AI for learning

The intro post of a five-part series on why and how inviting in student voices early has shaped the way I build, test, and think about AI for learning

Welcome to 29 Seconds of Farcaster, a quick recap of what you missed on Farcaster over the weekend, delivered to your inbox, every Monday. Brought to you in collaboration between @bradq and the GM Farcaster Network. What did you miss this weekend? 29 Seconds of Farcaster ... what did you miss this weekend? @jacob dropped Zora Creator Coins announcement, many big supporters jumped on board; @jessepollak shared his favorites https://farcaster.xyz/jessepollak/0xe515b362 V announced that we will ...