Tootfinder

I’ve been using Rails for close to 20 years. I use it for multiple active projects today. I rely on both its technical upkeep and the human infrastructure that makes that upkeep possible.
DHH needs to go. He is an albatross. He provides close to zero technical value at this point, and does a hell of a lot of human harm. Worse than deadweight: he’s a liability.

Brutal similes
Using #AI is an ethical choice.
I know that there cases when an #LLM could make my job easier. Which doesn't mean I'll use one. Just like I won't be buying cheap junk gadgets that could help me with some random stuff a bunch of times before they'll end up on a trash pile.
Yes, sometimes I am curious what an LLM could come up with. But then, there are people who are curious how many donuts they can eat before throwing up. A waste of good donuts.
What world would you rather live in? One where you put a little more effort in your job? Or one where LLM helps with with your job, but you can't enjoy your free time anymore because the capitalists are using LLMs to turn every single aspect of your life into a nightmare, and eventually your employer just makes you do more and more until you're thrown out? But at least you will get a monthly trial of a statistical "friend" to "talk" about your trouble to.
Yeah, you can claim that training models does the most harm, and that's already happened, so not using them doesn't change much, and all the energy spent on it would be wasted. Or use the traditional "others" fallacy — others will use it anyway, others will fuel the vicious circle, so why renounce convenience. It's like when you learn that your dinner is human meat, and you decide to eat it anyway, because not eating it won't bring that human back to life, and if it's wasted, then their death will be for naught.
#AntiCapitalism

CORRECT: COndensed eRror RECognition via knowledge Transfer in multi-agent systems
Yifan Yu, Moyan Li, Shaoyuan Xu, Jinmiao Fu, Xinhai Hou, Fan Lai, Bryan Wang
https://arxiv.org/abs/2509.24088

CORRECT: COndensed eRror RECognition via knowledge Transfer in multi-agent systems
Multi-agent systems (MAS) are increasingly capable of tackling complex real-world tasks, yet their reliance on inter-agent coordination, tool use, and long-horizon reasoning makes error recognition particularly challenging. Minor errors can propagate across agents, escalating into task failures while producing long, intertwined execution trajectories that impose significant costs for both human developers and automated systems to debug and analyze. Our key insight is that, despite surface diffe…

Yesterday I finally managed to close my account on #Vercel . I was hosting some stuff for free, so I was costing them money instead of giving them anything, but I still don't want to depend on them in any way.
But to the point, I just checked the online conversation about them... and I'm sad to notice how quickly people¹ forgot that Guillermo Rauch was very proud of sharing time with Benjamin #Netanyahu. He was so proud that he decided to share that moment with the entire world, so he showed us how little he cares for human rights, dignity and decency.
That's why I deemed necessary to remind folks about this serious contention point and inspire them to look for alternatives.
1. Including people who seem to care about these topics, which makes me think that the information didn't reach them in the first place. Let's help them know.
#boycott_vercel #BoycottIsrael #FreePalestine