yes, you can read the logs, and see full actions, the agents can send messages to each other, and I am also about to make it possible for them to optionally read each other's traces

Yeah, I guess we have sandboxes qwith our various code environemnts, and then we've been developing programs that run agents to do various things, which we iterated on to make them better at the task.

For example, one spawns copies of the env with a PR and runs agents in the dev env to verify by running and demonstrating functionality and then comments on github

another one is just a generic software factory that spawns a bunch of agents to coordinate on some repo, others do a redteaming flow, etc...

What do you do with those agents? It's useful if you want to iterate on a flow and have more control over the orchestration/environment

hey HN! happy to answer any questions

this kind of tool is especially useful in longer running tasks to enforce your intent without having to check in on your agent all the time

Can users stack Quibblers, so Quibbler 2 corrects Quibbler 1 if, say, it fabricates an issue in the code it's reviewing? If so, have you found an optimum number of Quibblers for the Quibbler stack? Also, might users form a Quibbler council such that multiple Quibblers review the same thing and form a consensus before proceeding?

I love the pixel-perfect precision with which this comment is straddling the Poe's Law line.

That aside I also love the concept of Quibbler Council and I'd get a kick out of seeing it in action.

MoQs - Mixture of Quibblers? Would be convenient to have them run on dedicated FGPAs. Then they can facilitate near real-time quibbing at the network level across all packets.

http://uzpg.me - philosophy, tech, my life

Relevant: reverse engineering modular addition https://twitter.com/NeelNanda5/status/1559060507524403200

https://uzpg.me/blog.html -> personal blog on tech and sci-fi + what i'm learning

Making something you can charge requires a lot of extra software work, and like other people said it creates a responsibility on the creator's part.

I'm researching how we can use technology to improve the way we learn and think.

I'm currently exploring how AI/Natural Language Processing can automatically organize personal knowledge and notes by tagging and linking ideas / content together [0], removing the friction of organization to help you find new connections and ideas.

If you're interested in this domain, I'd love to talk to you!

[0]: https://github.com/Uzay-G/espial

Couple months ago karlicoss the developer of Promnesia posted in the collab post https://news.ycombinator.com/item?id=29760122

I made a discord server in response (we mostly idle) to discuss knowledge management tooling

Awesome Knowledge Management https://discord.gg/XPNeDSQE2j

Would love to continue to chat about existing tools and new ideas in this space!

Nice! Archivy [0] has some of the same goals with integration of web content, also using Readability. Instead of going the highlight route, each article becomes a note in your knowledge base you can edit / add to.

It's really exciting to open source a project, Lurnby looks cool! What do you use the spaced repetition for?

[0]: https://archivy.github.io

Woah. Very cool. You seem to make it really easy for people to get started and tick a lot of the boxes around privacy.

Today's been a bit of a nervous day for me actually haha, I didn't really expect this attention. Open sourcing just felt like the right thing to do for this app and because I just don't have all the skills :D

The spaced repetition at the moment is for the highlights. All highlights get marked at an initial level 0 and then move up or down depending on you reviewing them.

Eventually gets to the point where a highlight is shown to you yearly.

In the future I was thinking of doing some spacing around finished books or articles. Things like - you finished reading X 10 days ago, what do you remember?

But don't think I've thought through the details around that enough.