Define obviously validation? What is the signal that tells you one is reasonable vs another?
I find the only way to do that is to look at it, if it passes some visual tests, try it, and then a/b test if it's any better than without it.
2,136 karma · joined April 8, 2011
I find the only way to do that is to look at it, if it passes some visual tests, try it, and then a/b test if it's any better than without it.
I think one of the particulars about Go is how hard the player base took it. Far harder than chess did. Far harder than Starcraft did (although arguably, Alphastar wasn't even that good strategy wise, it was just better mechanically even with preventions. No one has adopted almost any of Alphastar's strategy)
Lee Sedol in particular was crushed by the experience.
Others found optimism and opportunity in it.
I don't think extrapolating the Go experience is all that useful across the board, although it does have some value, and perspective, and it was a fantastic article I enjoyed reading.
Games have cheating, because cheating is easier than getting better.
Before AI, there was rampant cheating. In Magic the gathering, it's shuffle cheating, or holding out cards, or whatever.
The ease at which you can cheat makes more cheaters. If you can get away with it, or if it's like Go, or Chess AI, it's trivial to do, and easy to not get caught.
Same with map hacking in Starcraft.
I don't know. I don't have any fully formed thoughts here, except that I think extrapolating the experience in this way is vastly overstating it's generalized impacts.
But I also could be very wrong. We are talking about predictions. No one can predict anything.
Predictions say more about you, and your perspective, than they do about reality.
But great read, enjoyed thinking about it all.
And I wish I could make even more deterministic. Maybe I can, but it can also be a bit challenging to sort.
https://github.com/notque/vexjoy-agent
I would prefer that be deterministic though. This thread has me considering what if anything I can do to make it forced. Like, I could do it with hooks, but that's not elegant at all.
But I don't expect anyone to every use my stuff. It's complicated as hell. But it's for me, and it works without me having to remotely think about the complexity.
I love that.
I only make it for me, so it's a bit complex and targeted towards me, and what I do, but it's pretty easy to adjust things.
https://github.com/notque/vexjoy-agent
Working on reading through Agent Skills, it seems we've converged on a lot of the same points, and I've never seen it, so trying to get an understanding of it.
Edit 1: I don't like all the commands. I just rely on a single router to automatically decide what I want, and that feels like the most reasonable way to me to communicate with it.
I don't want to remember things. And that's the way for me to scale the number of skills and activities. I don't have to think about them.
Edit 2: We have very different routers.
https://github.com/addyosmani/agent-skills/blob/f504276d8e07...
vs
https://github.com/notque/vexjoy-agent/blob/main/skills/do/S...
I personally wouldn't call theirs an intelligent router. They are dancing between a few different skills. We have extremely different setups there.
But of course, I'm using way more context to get it done. I'm even sending it out to Haiku to build the route choices.
I choose to use tokens to make things better for myself, not everyone would make the same choice, so I certainly see why they are using a few skills, and composing them.
Edit 3: This is much easier for a user to wrap their head around because there's much less.
I am only focused on the best improvements I can make that show value for my use cases. This is straight foward to reason about.
This seems like a nice way to get the best concepts for people trying to understand them. I commend them for a clean, simple approach.
Edit 4: Yeah, I think there are some things I can learn from them which is always good.
I especially like simple decisions like collapsing the install details for each harness in the readme.
I'm going to read over the entire thing and look for opportunities to improve my stuff.
We are all working together, learning, testing, building, trying to find the best way to implement things.
A submarine cannot swim.
It's a departure for Anthropic releasing skills to us to use, versus gating them and hiding what they do.
It feels important to me how skills are changing to be hidden and charged more, but I haven't seen any other discussion on it.
He was paid to work on it. That stopped, he continued to work on it in the hopes he could find someone who would hire him to work on it.
That wasn’t true, no one has funded it.
So due to the economic system he no longer maintains it.
That’s your economic system at work. No one is pretending it isn’t there, this is the outcome of it
Why is it never the responsibility of the people using it?
If anyone cares enough they will. People didn’t care enough to pay, so maybe no one cares enough to fork and be the new unpaid custodian
The reason for something to exist is someone finds joy doing it. Especially when they are unpaid.
The sadness should be focused on his inability to support himself with a tool that clearly a lot of companies, and people are using and gaining value for.
If no one cared enough to support the project, why does anyone care enough now? It all sounds hollow. Nothing bitter about it.
When you work on a project, any project, you have a responsibility. At some point we all can stop, and become free to not have that responsibility.
The problem is the base harnesses don't call them aggressively enough. Not that they don't work.
Anything that can be deterministic, should be
I used up 1/3rd of my context in less than a day. I am working diligently to do whatever I can to lower token usage.
I am on xhigh.
So yes, for the same tasks, usage runs out faster (currently)
All of us doing crazy agentic stuff were fine on max before this. Now with Opus 4.7, we're no longer fine, and troubleshooting, and working through options.
I've been playing around with doing that with a cron job for a "dream" sequence.
I really want to get them out of main context asap, and where they belong, into skills.
it runs indefinitely without my feedback or until rates limit it.
The AI agents built all of it. Not me. I realized I hadn't looked at it in awhile, and was looking for new ideas to add to my own ai system i built after learning from it.
So I thought I'd share it incase it was interesting to anyone else.
Especially the people who can't get claude code to not stop for them. I have been the entire time able to have it run indefinitely without my interaction. :)
I think i started it with a single prompt, and then told it to build everything else out it needed. That it's in control. I didn't write any of what is in the repo. I didn't tell it what to do.
It's just what it did.
EDIT: also i should commit what it has... it doesn't care about committing, so there's a whole lot of stuff it's built that isn't up there. Let me do that now.
EDIT 2: also, nothing is my ideas or written by me. It's all the collective. So please don't think i made the readme, or anything. I am not communicating ideas here, I let them do their own thing
I have a custom learning system. We are all trying things, that's where ai development is.
None of us know the best solution. We are all exploring in paths. I don't find memory and persistent long term context to be an issue for me, but I am using a full custom ai claude code setup, so perhaps I have sorted it for myself. Unsure.
Can you give a specific example? Like, talk through your workflow so I can understand it better?
i have checks of which types of repos i'm in with branching dev flows for each one.
it's going to be hard to communicate all of this genericly, but i am trying.
there's a lot of patterns i think are helpful for me.