Preserve context in the lead chat - let the subagents fill up their own contexts then only return the necessary information.

Why not just have an agent that can branch its context?

Forking conversations have been a thing for a long time, and it's not the same thing (e.g. you may have 50% of your context used up at fork time that is carried into "both agents" afterwards)

Also sometimes _not_ having the context is what you want.

Eg when I have the AI do a self-code-review before bothering a human, I also want to the AI to give the draft PR to a sub-agent that only has the publicly available context that a viewer of the final PR would have; and not all the accumulated reasoning that lead to the writing of the code in the PR.

It is exactly the same, you’re just doing it wrong.

So in your mind forking a repo removes the commits from one of the branches?

Why hire a junior developer if you have a perfectly competent senior developer already on your team?

You can do a code review on a "less capable" model that costs less, and the key model gets its output / summary, then you can have that model build a plan, and feed it to cheaper models. It's a more efficient approach than just running everything through Opus, and now that Sonnet is a lot better I'll probably use them more frequently, one thing to note is don't ask it to spin up endless subagents, I'd cap it to 2 or 3 at a time, otherwise, yeah you'll hit your limit extremely quickly.

Code reviews also work better in sub agents because the reviewer agent didn’t write the code being reviewed.

(More precisely, doesn't have the reasoning from when the code was written)

here's a simple task. Collect all info from unstructured email signatures, 3 years of small business emails.

Opus directs and starts haiku/sonnet subagents. Much more efficient than opus reading it all.

Without this explicit direction, Opus burned through usage to create complicated regex parsers & 8 excel sheets.

Because two agents are faster than one.

The models get dumb as context fills. Subagents allow them to accomplish a task with minimal context rot. You can also use cheaper models for subagent tasks

Money is money too. Increasing contexts costs non zero money, even with cache hits. Also context rot is a problem that subagents help with

> Time is money. Parallelism is very helpful optimising one to get the other.

Parallelism is fantastic when it actually speeds up the entire pipeline, but in my experience most people's jobs (at least the ones for which AI is currently relevant) involve a lot of overlapping "hurry up and wait" branches that drastically blunt the real benefits of that sort of parallelism.

There may be specific situations where it makes sense to do it, but just immediately going full gastown on anything AI related seems like such a giant waste to me, of both money and finite world resources.

This truism is intuitive to everyone but always funny to me how everyone never has any time, needs to save time, needs to hire staff workers for every mundane job and robots can't come soon enough… all so we can binge watch Game of Thrones and 90 day Fiancé.

And watch 10 hours of football on Sunday for our DraftKings bets.

Money is also money, which anyone who makes heavy use of parallel subagents will quickly learn.