Care to elaborate

Even with the strongest guardrails, no llm can be left unsupervised. And believing an llm can supervise itself is madness. For every agent you deploy as a “team” you will have multiplied diminishing returns.

I’ve iterated the process hundreds of times and even with a strong specification and tests, an llm can meet the spec and pass all the tests and still build the wrong thing.

This multi-agent workflow idea is worse than web3.

But it’s a YC thing so someone will make money.

I also don't believe in this multi-agent swarm thing. I just use Claude Code "manually" and this is still useful for that.

definitely understand the perspective. I feel like theres some dev work that completely requires me to be single threaded and just think hard, but there other modes where i can jump into whack-a-mole mode and try to work on a bunch of github issues in parallel, or sentry issues, or little ui bugs and the thinking goes more into reviewing plans and prs and i dont have to be as attentive to each individual session

There are individual tasks that can be “swarmed”.

Constructing software requires supervision and a continuous feedback loop from a discerning human.

agreed