Today’s trajectory toward chain abstraction and interop points to this: soon no one will care about where your application is deployed. As long as it can interop, as long as a user base can access it, and as long as it can be secured by enough operators. The doors will swing wide

Web3 projects have normalized the most egregious funding and incentive scheme - the app-specific token. I can’t fathom a mature web3 keeping this around. Maybe a few more years before the world gets tired of getting screwed.

The oracle problem is much better framed as a truth-finding problem

FuelVM and sway are grossly underrated.

gpt 4o is till a goated model. My daily driver. Sometimes more concise than o3.

I am quite excited to finally announce a theoretical framework for Veritas! I hope this interests some people and I encourage you to take a look!

I have observed something interesting during the recent launch of Veritas . The truth-seeking methods used in Veritas are VERY similar to the revolution we’re seeing in decentralized ML. Humanity’s collective intelligence will provide the highest signal used for this

A recent article by michaellwy illustrates a huge problem in current prediction market design. Veritas presents a solution - a breakthrough in resolution mechanisms.

The deeper I go into Veritas the more I understand it, but the more confusing it gets. Really hope I’m able to communicate it in an intuitive way

From time to time, I’ll share early-stage projects that are still developing but show strong potential. Todays guest is Veritas Misinformation travels faster than facts, and centralized sources of authority are losing credibility. Veritas offers a different approach to

Regex and Markdown don't mix. LLMs love both, but consistently struggle to generate content when combining both standards. We need a unifying standard.

Released a deeper dive into Veritas 's truth-seeking mechanisms. I think it's a pretty novel InfoFi game that I hope will be able to steer humanity towards efficient coordination. Take a look here: veritas.discourse.group/t/refining-ver…

I’ve been interested in experiments in credit-money recently. Trying out circles! Circles app.metri.xyz/p/0xaf2f46BF80… x.com/aboutcircles/s…

milian TEEs isolate, but that's the point: they provide a hardware-enforced confidentiality boundary. Even w centralized production, we can dramatically improve trust assumptions via on-chain attestation, role permissioning, and kms for off-enclave data sealing.

We shouldn’t be having to send out messages asking information providers to “wait to propose!”. This is a fundamental problem with using an optimistic oracle.

The world truly does need Ethereum. Ethereum is slow and bulky because that’s what you get when you design the most decentralized, censorship resistant POS consensus protocol in the world. The faster you get the less nodes you can support and the less sovereign your network