Trick or treat? p(doom) to maximum! Slightly Unaligned Slopcore Tutorial: a frenetic #slopcore video with just the right amount of AGI goosebumps for Halloween 🎃 No video models used, all sprites and code! Claudia is an open character by @anabology (claudia.gallery)

AI is rough if you're a software engineer who loves the engineering more than the software. It’s incredible for those that care more about the end result than how it was built. Make the shift.

What if vibe coders are just entry-level AI engineers? The more they interact with AI and engineers, the more they learn, eventually becoming skilled at knowing what’s under the hood. In our team, several people who started as managers or designers use AI daily to make small personal apps or prototypes to communicate better with engineers. They ask questions and learn non-stop, and their prototypes grow more elaborate and solid as we give feedback. This is a new time where careers blur and people can learn faster than ever if they’re curious and smart. Does this match your experience?

Imagine we could run this locally in a pair of AR glasses and connect it with our databases and tools. It would entirely redefine computing as we know it: a 100% personalized computing experience.

Ten minutes, four AI tools, one Warhol image. Each one remade it its own way, and none of them does it the same twice. How do you keep a brand under control when everyone produces with AI?

Hot take: line-by-line code review won't survive agentic engineering. It can't keep up with agent PR volume. What scales: strong types, deterministic linters, QA automation, proofs. We skipped those guardrails because they were expensive. Agents just made them cheap.

¿Y si el valor de un ingeniero ya no estuviera en escribir cada línea de código? Javier Toledo, cofundador y CTO de The Agile Monkeys, analiza en esta #entrevista cómo la #IA está cambiando el desarrollo y qué habilidades ganan peso. Toma nota. bit.ly/4ij4NcV

This has been a common topic in our team lately. You’re likely splitting the work too much if you keep using Opus 5.5 the way you used Opus 5.

I truly don’t get why some open-source communities reject AI-assisted work. AI could be the best thing to happen to open source ever. It lets us automate solving most bugs and create new features in the background while we work on other things at our paid jobs. I get that some people want to keep coding by hand as a hobby, but for real-world use, value delivery is what counts, and no hand-run project can deliver the same or more value than one using AI. My prediction is that projects that forbid AI contributions will be forked or forgotten in favor of those that allow them.

For some people, coding was the only valuable skill they had. If you made your money by selling your ability to write JavaScript, CSS, Python, or whatever else, I understand why you are pissed now. "Writing code" is worth very little today. The sad part is that learning to code wasn't easy. It takes years for someone to become a skilled programmer. Imagine watching all of those years go down the drain. If you are a programmer (I mean, just someone who knows how to write code in a particular programming language), your days are numbered. If you still think AI can't do your job, you are just in denial. But honestly, writing code was never that valuable anyway. If you were a code pusher, you already knew that the best people at your company weren't just writing code. They came up with new ideas, designed solutions, architected whole systems, knew how to troubleshoot and fix problems, found different ways to build better software, and, in general, spent almost no time writing code. Not only are those people thriving today, but they're also the ones who will ride the AI wave and learn to bend it to their will. So don't cling too long to what you lost; embrace what's already here. AI is a tide that lifts all boats, but you have to be open to it.

Today, the bottleneck is you. So we built Shepherdr: an open source queue for every coding agent on every machine, on top of @herdrdev. The point isn't the tool. It's that every engineer can now build their own.

So true

Deterministic validation, QA automation, strong type systems, and proofs can help to eventually eliminate code reviews and remove another bottleneck in software development automation. These guardrails were too expensive for most projects before, but are easy work for a coding agent.

Claude one-shot a making-of video of the INSANE infomercial. I think it's actually more interesting than the ad itself. I forgot how much iteration I went through to get this one 😅

I wanted to push @ElevenLabs v4 to the limit for explainer videos. Plugged it into my #slopcore pipeline and… this came out: An 🔥INSANE🔥 infomercial for Shepherdr, a product I never meant to sell, with a cameo from @anabology's @claudia_popstar and the goats 🐐.

This is why I firmly believe most consumer software will be open source. Code is becoming just a cache of what agents can generate anytime. Limiting access incentivizes consumers to ask an agent to reverse-engineer it and create a free clone. Open-source alternatives will also significantly improve in quality thanks to AI (as we’re already seeing in projects like @omarchy), so we likely won’t even need to reverse-engineer anything.

Be the G.O.A.T. ...or farm the goats! 🐐🐐🐐

If I’ve learned one thing from working on SDLC automation, it’s that AI doesn’t just change how we write code. It forces us to rethink the entire software development process, including how we structure engineering teams. I once heard @elonmusk say that “the most common mistake of smart engineers is to optimize a thing that should not exist.” I think this applies perfectly to SDLC automation. It’s easy to fall into the trap of automating processes that were designed around human limitations instead of questioning whether those processes should exist at all. We’re still figuring out what the future looks like, but one thing is becoming clear: our job as engineers is shifting from writing and reviewing code to building the pipelines that allow AI to work autonomously, reliably, and securely. We’re even questioning whether PR reviews, or PRs themselves, still make sense. After all, PRs were largely a coordination mechanism designed for humans. What matters increasingly is having comprehensive tests that catch regressions, automated QA that exercises the entire UI, and rigorous architecture and security validations. The future of software engineering isn’t about reviewing every line of code AI produces. It’s about engineering the systems that make its output trustworthy.