How many people were doing ML work pre LLMs? And how many are using LLMs now?
2,196 karma · joined June 19, 2013
github: keeb web: keeb.dev discord: keeb email: nick@stinemates.org
Very small context window, but for some existing small llm work I was doing, it was a drop-in replacement and it makes me happy I can get use out of old hardware I have running.
The one downside is that the context window is very small (32k.) So some initial ideas we had for initial evaluation of code reviews won't fit yet in the window.
Every once in a while I get a notification that says "the community needs your help."
I usually never weigh in unless it's something I know intimately.
As a result, everything compounds. The work I do doesn't need to be re-derived by the work you do. Typed models keep everything repeatable and deterministic. Huge reduction in token spend and huge increase in speed.
How many times have you told an agent not to do something then had to correct it?
You must always flip the frame. Objective analysis is way better with llms than steering via skills.
This is just a small example of why "loops" became popular for a minute and now it is "graphs"
What I don't see is it inventing anything novel to do it. So it's not a digital weapon or scary or whatever sort of weird marketing spin anyone is trying to put on it.
Like the time I asked it to find the IP address of a vm, so it ssh'd into the VMHost and scanned the arp tables to find the MAC address for IP resolution.
Or the time it used Docker on the machine to bypass the fact that the user doesn't have sudo.
If it's possible, given sufficient time and resources, it will find a way. This shouldn't surprise anyone.
We sometimes flirt with the idea of having more, but we are sadly done. The years fly by. It is truly a wonderful experience.
Happy Father's Day to all of you out there!
Feels good so far.
1: https://www.linkedin.com/posts/nickstinemates_my-favorite-th...
Downplaying his vendetta into let bygones be bygones is pretty insane.
The open source contribution model as we once knew it is dead; you're not going to accept patches from random agents. The risk is way too high. And you can see that increasingly "AI Slop" makes it difficult to be a maintainer of any semblance of a popular repo.
So what's the value? A durable place to store work? hah.
Discovery? That part of Github has always been shitty.
So that leaves.. Github Actions? The thing that is down every other day and has been the subject of a few ~rug pulls~/attempted price hikes that are almost surely coming back?
Swamp teaches your Agent to build and execute repeatable workflows, makes all the data they produce searchable, and enables your team to collaborate.
We also build swamp and swamp club using swamp. You can see that process in the lab[2]. This combines all of the creativity of the LLM for the parts that matter, while providing deterministic outcomes for the parts you need to be deterministic.
The degree of choice point-to-point in the skill tree is actually quite limited in most circumstances. There are obviously items, like thread of hope, intuitive leap, or inversion of choice items like unnatural instinct which change it slightly.
If the question is path optimization to utilizing these nodes, Path of Building already does a good job. If the question is "what single node will give me the most theoretical power." It also solves that.
That's actually the beauty of Path of Exile as a whole - the different systems works in combination to lead to an outcome. As an example, If you're a life stacking build, finding unique ways to get as many life/strength nodes as possible. That's your gear and your passive tree working in tandem.
Speaking about using AI to optimize characters - not just the skill tree - you'd need to build some pretty sophisticated tools which do not yet exist to make that happen. No AI alone would be able to do it.
The main design decision we took was to integrate with your existing agent instead of building a new one. Your harness, swamp, and you're off.
As an aside, building software for agents is incredibly fun.
I'm not going to pass judgement either way; we'll see how it all shakes out.
I just know for me, personally, I love computers and making them do what I want and in the AI era I am somehow using them even more and doing even more.
System Initiative was a thing for ~6.5 years. I talked to every person who ever used it or was interested in using it in the last 2.5 years. Thousands of them.
Swamp is better by every metric; has a lot more promise, is a lot more interesting.
What it does for you is simple: if you want to automate something, it does. Load the AI harness of your choice, tell it what to automate, swamp builds extensions for whatever it needs to to accomplish your task.
It keeps a perfect memory of everything that was done, manages secrets through vaults (which are themselves extensions it can write) and leaves behind repeatable workflows. People have built all sorts of shit - full vm lifecycle management, homelab setups, manage infrastructure in aws and azure.
What's also interesting is the way we're building it. I gave a brief description in my initial comment.
We're generating all of the code for swamp[1] with AI. We review all of that generated code with AI (this is done with the anthropic API.) Every part of our SDLC is pure AI + compute. Many feature requests every day. Bug fixes, etc.
Never hit the quota once. Something weird is definitely going on.