Intent-based systems aren't fragmented - they're just apps - their solver systems are. Why? Because different intent apps introduce: - Different commitment systems (intent expressions) - Bespoke settlement systems (verification conditions) - In the case of cross-chain,

It's concerning that the pendulum has swung from fat protocols all the way back to fat apps, "starved and money only" protocols. I strongly disagree a starved protocol's native token can be a long-term store of value, and this includes Bitcoin. Declining incentives to keep the

The Rise of Attestation Based Insurance Protocols I have already wondered why insurance protocols haven't grown to become a major category for Defi despite its prominence in traditional finance. Recently I’ve come to realize why that's the case, and more importantly, why

This is a great article by Arjun | LI.FI, but he could have saved 3000 words and just shared a single diagram

The ongoing debates about "layer x vs layer y" is missing a crucial point. We should shift our focus to assets and their authoritative clearing runtime: - USDC's L1 is Circle - USDT's L1 is Tether - any blockchain is L1 for their native and natively issued assets - any

Once this is in place - intention is *all we need*. Because: intents can also be represented as valuable state.

Maybe not “this cycle”, but I think programmable attestations are going to vastly expand crypto’s tam smart proofs are the new smart contracts

off-chain verifiable but not necessarily decentralized apps (vApps?) - stateful - solver networks - builder networks - TEE Apps - AVSes - stateless - co-processors - provers - relayer networks many off-chain apps are not inherently verifiable: -

we're too arrogant thinking crypto x ai is about us building a more (insert your favorite adj. here) infrastructure for AI. it's more likely going to end up being an un-censorable app store for sovereign AI to build apps for humans living on UBI checks, with a token that

If we were able to drop the barrier of launching a solver to as low as clicking a button, solver reputation system would not be necessary, because every service provider, either crypto native or in real life would have their own solvers, and everyone just transacts through their