big fan of the 'respects my hardware' philosophy. i feel like 90% of ai tools right now are just expensive middleware for openai, so seeing something that actually leverages local compute (and doesn't leak data) is refreshing

reminds me of twitch plays pokemon but for an llm.

honestly surprised anthropic hasn't nuked the api key yet given what internet strangers typically do with unsecured input fields lol

if i have to estimate something > 2 days i just assume im guessing.

usually means there's hidden complexity i haven't found yet. i estimate until the subtasks are small enough (like 4h chunks), otherwise its all just feeling based numbers.