Korede (korede)

Probably a nice topic, but EAS (expo.dev) of the best software products I've ever used. The amount of complexity that is abstracted away from cross-platform mobile app deployment is immense.

• 1 reply
• 0 recasts
• 1 reaction

We should only start getting scared of dev AI tools when companies start training models from first principles a la AlphaZero (i.e. handing them the docs and keeping them running until they figure out *everything*). Depending on existing code will continue to produce the limitations of existing code. But maybe this is only possible in a post-quantum world.

• 1 reply
• 0 recasts
• 0 reactions

While "vibe coding" is fun for rapid prototyping (especially on frontend-heavy projects), it seems that an under-appreciated use of LLMs might be to have architecture-level discussions about complex problems, making conceptualizing solutions and creating design documents way faster.

• 1 reply
• 0 recasts
• 1 reaction

I think for many people it’s probably the (perceived) added complexity. Another thing is the “perf overhead” with type validation per-request. Personally, I’ve used it on almost every serious project for 5ish years now and it’s been a huge help.

• 1 reply
• 0 recasts
• 1 reaction

We’re progressively moving from ethers to viem at Kwenta https://github.com/Kwenta/kwenta

• 0 replies
• 0 recasts
• 1 reaction

George Hotz Chris Lattner

• 0 replies
• 0 recasts
• 2 reactions

L3s are usually application-specific chains (have custom features not available on the EVM), but still rely on L1 for security and L2 for scalability. Best example I can think of is to imagine if the dydx chain was built on the OP stack instead (with the onchain order book on the L3)

• 1 reply
• 0 recasts
• 1 reaction