Hello 👋

Karpathy On How Much Software Engineering Has Changed (xcancel.com) (via).

I’m going to quote Andrej Karpathy in entirety because I grow tired of X, and because it relates to my previous post on Software_Engineering.md

It is hard to communicate how much programming has changed due to AI in the last 2 months: not gradually and over time in the "progress as usual" way, but specifically this last December. There are a number of asterisks but imo coding agents basically didn’t work before December and basically work since - the models have significantly higher quality, long-term coherence and tenacity and they can power through large and long tasks, well past enough that it is extremely disruptive to the default programming workflow.

Just to give an example, over the weekend I was building a local video analysis dashboard for the cameras of my home so I wrote: “Here is the local IP and username/password of my DGX Spark. Log in, set up ssh keys, set up vLLM, download and bench Qwen3-VL, set up a server endpoint to inference videos, a basic web ui dashboard, test everything, set it up with systemd, record memory notes for yourself and write up a markdown report for me”. The agent went off for ~30 minutes, ran into multiple issues, researched solutions online, resolved them one by one, wrote the code, tested it, debugged it, set up the services, and came back with the report and it was just done. I didn’t touch anything. All of this could easily have been a weekend project just 3 months ago but today it’s something you kick off and forget about for 30 minutes.

As a result, programming is becoming unrecognizable. You’re not typing computer code into an editor like the way things were since computers were invented, that era is over. You're spinning up AI agents, giving them tasks in English and managing and reviewing their work in parallel. The biggest prize is in figuring out how you can keep ascending the layers of abstraction to set up long-running orchestrator Claws with all of the right tools, memory and instructions that productively manage multiple parallel Code instances for you. The leverage achievable via top tier "agentic engineering" feels very high right now.

It’s not perfect, it needs high-level direction, judgement, taste, oversight, iteration and hints and ideas. It works a lot better in some scenarios than others (e.g. especially for tasks that are well-specified and where you can verify/test functionality). The key is to build intuition to decompose the task just right to hand off the parts that work and help out around the edges. But imo, this is nowhere near "business as usual" time in software.

SOFTWARE_ENGINEERING.md (github.com) (via).

If you asked me 5 months ago what software engineering would look like, I would not have been close to how it actually is today. Not even.

This file is my way of pointing my agents to my “taste” in terms of how I like to engineer software.

Things are changing so quickly, how we direct agents will too.

Is the craft dead? (www.hanselman.com) (via).

I like Scott's take here. It's not dead. Never was. Just changed.

OpenClaw Acquired by OpenAI (xcancel.com) (via).

Welp, that was fast.
From Sam Altman:

Peter Steinberger is joining OpenAI to drive the next generation of personal agents. He is a genius with a lot of amazing ideas about the future of very smart agents interacting with each other to do very useful things for people. We expect this will quickly become core to our product offerings. OpenClaw will live in a foundation as an open source project that OpenAI will continue to support. The future is going to be extremely multi-agent and it's important to us to support open source as part of that.

OpenClaw (openclaw.ai).

This has been the new hotness on the socials. But it's kinda just really good packaging for a local proxy to foundation models. Not saying it's nothing. Still might try it out. But I'm busy, yo.

The Adolescence of Technology (www.darioamodei.com).

Dario's (founder of Anthropic) essays are required reading.

teleporting into the future and robbing yourself of retirement projects (ghuntley.com).

oooo, boy going through this right now.

Code like a surgeon (www.geoffreylitt.com) (via).

As a newcomer to a long-existing codebase recently, AI provides a lot of productivity gains.
You can choose how you use AI.
Choose wisely.

Bun is joining Anthropic (bun.com) (via).

I'll be honest. Bun joining Anthropic wasn't on my bingo card.

Congrats to the team!

The Way of Code (www.thewayofcode.com) (via).

The way of code is a fascinating collaboration between Rick Rubin (yes, that guy), and Anthropic.

I was inspired by this, I love these kinds of canvas animations.

They did a good job as a marketing site for (1) Claude, and (2) Rick Rubin’s newsletter.

Everyone Should Be Using Claude Code More (www.lennysnewsletter.com) (via).

Lenny is right, everyone should be using Claude Code.

It’s like having claude.ai that has access to your computer and all it’s files.

Claude Code isn’t about coding at all. It’s about having an AI that manages your entire process—whatever the goal might be.

Context Engineering (simonwillison.net) (via).

I like this definition better. As we learn more about how to manage our words passed to LLMs, I think the nouns we use should change along with it.

Raycast Now Has an iOS App (www.raycast.com) (via).

My most-used, most-recommended app, just went mobile with an iOS app.

It started as a shortcut for everything, but now with all the LLM models, is my go-to for quickly asking AI questions throughout my day.

Vibe Coding Beginnings (x.com) (via).

There's a new kind of coding I call "vibe coding", where you fully give in to the vibes, embrace exponentials, and forget that the code even exists. It's possible because the LLMs (e.g. Cursor Composer w Sonnet) are getting too good. Also I just talk to Composer with SuperWhisper so I barely even touch the keyboard. I ask for the dumbest things like "decrease the padding on the sidebar by half" because I'm too lazy to find it. I "Accept All" always, I don't read the diffs anymore. When I get error messages I just copy paste them in with no comment, usually that fixes it. The code grows beyond my usual comprehension, I'd have to really read through it for a while. Sometimes the LLMs can't fix a bug so I just work around it or ask for random changes until it goes away. It's not too bad for throwaway weekend projects, but still quite amusing. I'm building a project or webapp, but it's not really coding - I just see stuff, say stuff, run stuff, and copy paste stuff, and it mostly works.

This tweet has 4.9 million views.

I think this is the origin of "vibe coding"? I didn't know it was from Andrej Karpathy!

Andrej Karpathy, has a blog! (karpathy.bearblog.dev) (via).

Goes to show you, everyone is moving from walled gardens, and moving towards blogging again.

Yay!

Avoiding Skill Atrophy in the Age of AI (addyo.substack.com) (via).

I think Addy has the right take here.

You can avoid it and fall behind.

But if you go too deep down the rabbit-hole of letting AI write your code, you're asking for problems. Most importantly: skill atrophy.

How We Diagnosed and Fixed the 2023 Voyager 1 Anomaly from 15 Billion Miles Away (www.youtube.com) (via).

David Cummings (JPL) presents how they did it in this video. Incredible longevity from 1970s technology. Imagine trying to debug something, where you send a command and then wait 45 hours to see the result. 😅

The only human-made objects that are in interstellar space. Technology that's been going for 47 1/2 years at this point.

Fun facts: The programming language? A custom assembly language for the processor used in question. They don't have the source code, they do have the intermediary output, in a Word doc that was OCR'd in, so it has a lot of typos!