Agents Should Write Code, Not JSON(blog.sshh.io)
blog.sshh.io
Agents Should Write Code, Not JSON
https://blog.sshh.io/p/building-multi-agent-systems-part-c0c
4 comments
Youre right and there are some assumptions being made here around the agent having enough context to work on a task without interrupts (e.g. team review, asking questions, etc).
Typically human equivalent time is based on a single person given all the potential information they need up front (which is not today how a lot of work is done).
Typically human equivalent time is based on a single person given all the potential information they need up front (which is not today how a lot of work is done).
You're over doing the self promotion, perhaps you could review the guidelines, specificly:
> Please don't use HN primarily for promotion. It's ok to post your own stuff part of the time, but the primary use of the site should be for curiosity.
https://news.ycombinator.com/newsguidelines.html
Show us things you're curious about!
> Please don't use HN primarily for promotion. It's ok to post your own stuff part of the time, but the primary use of the site should be for curiosity.
https://news.ycombinator.com/newsguidelines.html
Show us things you're curious about!
ack thanks -- didn't realize this was against the guidelines
“ Long-horizon tasks: We are increasingly solving tasks that take hours of human equivalent time. Agents are now able to maintain capability even as the context window fills with thousands of tool calls. The human-equivalent time-horizon continues to grow”
Because the most efficient humans don’t do this, they have short cycle times. This “agents can operate without intervention” always felt like an anti-pattern to me.