Comment by zarzavat
6 days ago
> The "manage a fleet of massively parallelized agents" gets me uneasy too
It shouldn't. The agents are not good enough to be used in a fleet.
I have Claude. It's fine, but I'm pretty confident that my low usage of Claude would out-compete a fleet of agents, because it feels like there's an inverse correlation between the number of tokens you spend and the quality of the resulting code (more tokens = more code to review, more bad code slips through)
That's basically my finding as well. Agent wrangling is herding cats. Working normally but tapping Claude for the smallest possible thing (look this up, finish this psuedocode, grab an example of this) feels like it works better all around—faster, safer, far fewer tokens, results in work that the team understands, aides flow rather than adding constant context switching...
Maybe I'm wrong and the time will come to hang up my editor and go open an Italian restaurant or something. Until then I'm really inclined to believe my own eyes.