Dev After Code @devaftercode

The “how” never mattered as much as we thought. It was always about the “what.” The “how” just became so complex that we devs started mistaking parts of it for the “what.” AI is stripping that illusion away really fast. That’s what feels so unsettling

Luna is SO good for fine grained tasks. I often find myself going straight from the cheapest option (Luna) to almost frontier level (Sol), basically skipping the midtier (Terra), because it usually fails for me. So I end up spending tokens on Luna, then Terra, only to have to ask Sol anyway... Almost never have to use Astra atm.

It feels like we're about to witness the *fastest* destruction of jobs in history, both high and low skill. There are still real constraints like chips, energy, data centers, reliability (almost there)... and regulation and organizational inertia. But the adoption side has almost none of the physical brakes that slowed previous technological transitions. Feels scary at times.

@theo Must admit, that caught me off guard and I almost spit my drink all over my keyboard 😂

Today, on underrated solo builder things... Letting coding agents manage tasks locally in .md files. No MCPs, no boards, no skills. Just a few lines in agents.md and they manage tasks as we work. I barely open the backlog anymore. Ask what's left then prioritize and refine with your agents while they keep the tasks in sync.

the last 20% to get a product shipped and successful is as difficult as ever this is where 99% of programmers have always fallen off the first 80% is a lot easier and so we're seeing more abandoned projects than ever

One thing I'm really liking about building with AI is making it use the product before I ship. Go and ask it to find confusing flows, useless stuff, places where you're polishing way too much... You'll be surprised!! Now the feedback loop starts BEFORE users even show up! 🔥

"The worst they can say is no" is missing half of the equation. If all you do is ask, but don't put in any of the work upfront to validate the request, you're not going to get anywhere. Yes, be bold, be audacious, but do the fucking work first.

Anyone here using Gelato for posters/framed prints? Thinking of using it for the little project I'm building. Mainly curious about print/shipping quality, paper options (170/200/250gsm, matte, gloss...). My current idea is 200gsm. Any experience or caveats before I commit? 🙏

@thekitze Same! Every few months I end up trying to build some new system with calendar notifications, reminders, checklists... Trying to hack myself into doing things before they turn into a last minute emergency. Somehow, my brain always ends up winning 😅 It has its perks though!

The other day I saw my friend's niece getting frustrated arguing with ChatGPT over one tiny thing. Somehow that turned into a tiny product I'm building (and will ship soon). I've been building it with Luna. Crazy capable for how few tokens it burns. Seriously, just WOW. 🤯

The real players know exactly what this map is.

Skills, harnesses, MCPs... Way too many concepts. Most die in days because they solve problems the next LLM/agent version won't even have. Just play with agents! Pitch a fun idea, business model, app, or daily problem. Let the agent challenge your thoughts, refine and build!

Making a little book about LLMs for PHP, JS and Java devs. The goal isn't to turn anyone into an ML engineer. It's to build enough intuition and vocabulary so LLMs stop feeling like magic. Even building a tiny language model along the way. Still cooking!

OK. It's time to rethink this. I've spend the last several weeks working on a harness that tightly constrains the agents to work the way that I want them to work. I set up all kinds of gates, and tests, and tools, and protocols, and ... And while I was heads-down getting that to work, the agents got a LOT better. So much so that when I came up for air, the need for my harness was obviated. Indeed, the need for _any_ but the most liberal of harnesses may be obviated. Just how good these things have gotten blows me away. I have had long debates with grok and codex about the structure of systems -- as if they were senior engineers. They often disagree with me and have their own perspectives. I have, more than once, found myself agreeing with their views. I have not given up on constraints and tooling. Unit testing is still important. So is CRAP and Mutation testing. These tools still find bugs and offer useful constraints, though they can leave scars. However, the agents have gotten so good that I can now give one a very significant task with a few guidelines and it will faithfully implement it. I can walk away for 40 minutes and when I return it will be done. CRAP will be satisfied, Coverage will be high, and Mutation testing complete. The architecture will be clean, and the code will be very good. The end result may not behave perfectly, but it's so close that a couple of tweaks usually puts it into place. What does this mean going forward? I'm not sure. But I'm beginning to think that harnesses should not treat agents as components within a software design.

Related to the idiocy of "software factories" Just using one AI coding agent is good enough and it can spin up more agents if it needs to You don't need to build that yourself

@

57 years ago

0 0 0 0 0

@marclou And most people need a real income to live before their entrepreneurship succeeds! Having a job isn’t the opposite of being an entrepreneur. Sometimes it’s what gives you the runway to keep building. Being broke is not a prerequisite for being a founder. Stop romanticizing it!

IMO AI today is basically the "infinite monkeys eventually typing Shakespeare" theorem, but we've learned how to rig the odds so they stop producing garbage and get there much faster. We didn't eliminate randomness, we just learned to bias it. Thing is, is that intelligence?

Maybe most SaaS dies in a few years, or the web as we know it disappears, or... Guessing the future is fun... until it scares you. Don’t overthink it. Build solutions to problems you run into every day. It's a fantastic time to build! Time will tell, that part hasn't changed.