Matt Rice ⛺️

This is what zk + intents gets you. Get USDC on HyperEVM instantly.

👋 Solana

Okay, would love some economists to help me here (cc @BlkBoxEconomist). I assume GDP is generally measured in real terms (i.e. adjusted by inflation), so if AI makes many things cheaper (productivity improvement), then we would expect either more quality or quantity of those things to be consumed or for the extra surplus to be used elsewhere to increase production there. Let’s use software as an example. Software gets cheaper to make for a given level of quality and function. This means I get more, better software at the same cost OR the same software at a lower cost. This is deflation (see TVs over the last 20 yrs). If the economy buys the same software for less money, then the real GDP of the software industry stays the same, but there is surplus to go elsewhere. GDP goes up. If the economy buys more/better software for the same money, the inflation adjusted value has gone up because of the deflation. So GDP of the software industry has gone up. GDP goes up. The only scenario where the productivity gains are not leaked from GDP is if the sane software is purchased for less money, the surplus goes elsewhere, and that surplus is used to chase supply constrained goods (land would be an extreme example) where each marginal dollar of spending produces less marginal value than the last. Is this analysis right? What am I missing? @dwarkesh_sp and others seem quite convinced that the cheapness of the tokens (productivity gains) will work against extreme GDP growth.

AI is going the way of wikipedia. Everybody says it's unreliable until it isn't. Our research team is pushing the boundaries of how reliable and accurate AI systems can be. Soon, UMA will be able to provide faster, more objective, and (incredibly) cheap truth.