It's the little things
Maybe it's okay that we have agents now. The interesting hard part about programming, where you'd get in a flow state and stay there, keeping the problem in your head for hours, I think that time is gone. But I …
I’m a software engineer writing about building software with agents.
Maybe it's okay that we have agents now. The interesting hard part about programming, where you'd get in a flow state and stay there, keeping the problem in your head for hours, I think that time is gone. But I …
The problem with typing, is that it just looks like typing no matter what kind of work you're doing. If you're not looking at the person's computer screen you assume...
I had a simple idea awhile ago checks watch...apparently five years ago! What if I could loop a song from the CLI to stay in the groove when coding? I called it looper Originally it could only play local files, …
I've been using Bear for years. It's where all my thoughts go - meeting notes, code snippets, random ideas at 2am, links I'll definitely read "later" spoiler: I won't. But here's the thing: when I'm deep in a coding session …
I’m going to do more of the Show Your Work thing here on my blog. Buckle up. I really like automation. Probably a little too much. Add my love of README files, and you’ll be able to see where this …
It definitely makes the size much smaller - we can all move faster with agents.
The idea of a "pirate" and "architect" is interesting but boy are we going to burn a lot of tokens. Maybe that's ok for the speed gain?
From Jarred Sumner on the PR:
Blog post with details coming soon.
It passes Bun's pre-existing test suite on all platforms (and fixes several memory leaks and flaky tests), the binary size shrinks by 3 MB - 8 MB, the benchmarks are between neutral and faster - and most importantly, we now have compiler-assisted tools for catching & preventing memory bugs, which have costed the team an enormous amount of development & debugging time over the years.
The codebase is otherwise largely the same. The same architecture, the same data structures. Bun still uses few 3rd party libraries. No async rust.
To try this, run:
bun upgrade --canaryPlease do file issues if you run into any. If this thread gets crazy I will lock it.
Note:
- Still some optimization work to do before this lands in non-canary version.
- Still some cleanup work to do (which will come in a series of follow-up PRs)
Ambitious.
LLMs + a well thought out test suite they can verify against can go a long way.
Phish isn’t my jam. But Daft Punk is. This piece resonated with me. We’re all still figuring this out.
From Andrej Karpathy:
I packaged up the "autoresearch" project into a new self-contained minimal repo if people would like to play over the weekend. It's basically nanochat LLM training core stripped down to a single-GPU, one file version of 630 lines of code, then:
- the human iterates on the prompt (.md)
- the AI agent iterates on the training code (.py)
The goal is to engineer your agents to make the fastest research progress indefinitely and without any of your own involvement. In the image, every dot is a complete LLM training run that lasts exactly 5 minutes. The agent works in an autonomous loop on a git feature branch and accumulates git commits to the training script as it finds better settings (of lower validation loss by the end) of the neural network architecture, the optimizer, all the hyperparameters, etc. You can imagine comparing the research progress of different prompts, different agents, etc.
github.com/karpathy/autorese…
Part code, part sci-fi, and a pinch of psychosis :)
Been trying to think of ways to leverage this. There have been some amazing examples so far:
What else is autoresearch going to push forward?
Vibed a site around my notes/pics on my trip to the Code w/ Claude conference: https://program247365.github.io/code-with-claude-2026-sf-recap/
With the help of Claude code, I’ve been able to make blogging extremely frictionless for me now.
Within 20 minutes I had this feature coded and deployed.
Wild.
We had a breakthrough in AI with transformers.
Where’s our breakthrough in AI with alignment?
Simon commenting on Jeremy Scheff's post:
Yeah I feel this both ways I think. I'm not sure which camp I fall into.
I think it depends. For personal software, it's amazing, the things I knew I could build but would take FOREVER, can now be done fairly quickly.
So I get to enjoy more software abundance to satisfy my every whim.