Pete

“If someone 50 years ago planted a row of oaks or a chestnut tree on your plot of land, you have something that no amount of money or effort can replicate. The only way is to wait.” lucumr.pocoo.org/2026/3/20/some…

The amount of mis-attributed music on @TIDAL is astounding.

@Sabilappdev @jasonfried slop

@jasonfried True, wording like that isn’t random. Even demos are designed to guide perception.

@danielrachlin @jasonfried slop

@jasonfried Those subtle details really do matter. Even 'handcrafted' works. It's never accidental.

@tntsportsmex @StreamMaxLA @TNTLA no pueden estar en serio. Quién quiere ver todo menos el final de un partido? No pagamos para 90% de las pelis tampoco.

¡Con lluvia y todo, Sporting se entregó al máximo! 😵‍💫 Así se vivió a nivel de cancha el gol de Maxi Araújo que significó el 4-0 en el José Alvalade ante Bodo/Glimt. 🗣️ 📲 @StreamMaxLA: bit.ly/4fYZ4aQ 📺 @TNTLA #BenditaChampions | #TNTSportsEnHBOMax

still being mad about woke in 2026 is cringe.

The divorce explains a lot of the energy this guy brings to social media.

It's crazy to me that you can get a MacBook for less than most new iPhone models.

@wesbos @syntaxfm GH cloud agents, believe it or not.

What’s your remote agent story right now? Claude code on a box? Cursor cloud agents? Claw on a mini? I saw the new open code stuff launch last night. Doing a @syntaxfm on the state of this right now

@ElenoAM Tlalpan centro tiene el mismo problema. Echan concreto y dibujan líneas para que se vea casi como piedra, y termina viendose ridícula.

Los empedrados de San Angel hacen de esa colonia inaccesible; deberían de quitarlos.

One of the more fascinating results of "agentic engineering" is regular people building their own ideas. Dudes like Robert and John Ganz knocking out things all on their own is pretty fun to watch.

The plot thickens

@bygregorr @yarotheslav is this slop performance art? Because if yes, you're pretty funny. If instead the intent of this account is sincere, you're a fucking idiot.

@yarotheslav What if the core issue wasn't the tech stack, but our willingness to adapt and integrate new insights into existing codebases?

Ask Claude Code to apply the learnings from this thread in your Rails app. You're welcome discuss.rubyonrails.org/t/active-stora…

*trump saying 'everything's computer' voice*: everything's gambling

Loser energy is off the charts.

Bitcoiners when bitcoin rises: We're going to <multiple of current price>! Bitcoiners when bitcoin crashes: 1 btc is still 1 btc, so who cares what the usd value is?

@_Evan_Boyle Is there any public documentation on the difference in performance or results between using a model like Opus via gh-copilot (CLI or IDE) vs its native harness (CC)?

Includes support for changing reasoning effort from low/medium/high when you select the model. Enjoy!

Copilot CLI nows supports opus 4.6 My new daily driver. Give it a go!

Every new model or tool that comes out sparks wonder and excitement which turns into these posts. Three weeks later the same guy is already moving onto the next model or tool.

I hope everyone's enjoying the sloppification of everything.

@badlogicgames @steipete @GergelyOrosz I have oscillated between at times feeling like I don't review my code enough and other times feeling like I'm reviewing LLM-generated code too much. The idea that the impt skill is knowing what needs review or more involved human supervision and what doesn't resonates a lot.

They also talk about people management skills being a prerequisit to being good at agent management. I think that is true only to some extend (having both deep tech and mgmt experience myself). It comes down to trust and learning abilities. I can build up trust with a human coworker by observing their work output, giving feedback and observing how they integrate it, and can rely on them being generally able to learn from mistakes. They can also proactively come and ask for help. All without me needing to babysit them. In case of clankers, that trust is simply not there yet, as they still make too many absolutely idiotic mistakes (CVEs on clawdbot's tracker are evidence of that, no shade, shit happens). You can try to "teach" them by encoding lessons learned into resources that automatically go into the context, but there's only so much you can stuff into the context. You can further coerce them into producing better outputs with feedback loops like type checks, lints, tests, whatever. But these also have limits. Especially if you let your clanker write tests. Until that changes, and I think there are some technical limitations that make it unlikely that it will in the near future specifically wrt to context and how current long context capabilities are basically smoke and mirrors, there is still massive value in checking in more on your clankers output than you would with a human coworker. That does not mean you have to check and know every line of your software. But not looking at the code at all is also a recipe for disaster eventually. We have plenty of examples to demonstrate that. Famously, Claude Code is now basically 100% vibe coded. And I think it shows. The same is true for clawdbot (again, no shade, single person project, massive scope, all good, except when it comes to security). I think you are still kidding yourself if you think you just need to prompt good and your swarm of agents does the rest. With current SOTA models, the real art of "agentic coding" is to develop a sense for what you can trust the model with and what needs human intervention and supervsion. The problem with that is that model behaviour is not deterministic, intuition for one model doesn't transfer to other models, even within the same family or reasoing level, intuitions for one coding harness does not transfer to other harnesses, and so on. Typing out boilerplate was never the problem. Clankers are good at that. You can also trust them to type out variations of established patterns and algorithms for the most part. But software is still fucking hard. It gets harder if you trick yourself into thinking the clanker can take on the hard parts.

Recommended viewing. "What if you could do 600 commits per day and none of it was slop? This is what Peter claims is doing" I would be very surprised if @steipete actually claimed that :D youtu.be/8lF7HmQ_RgY?si…