Gioele SL Fierro

@adahstwt Super cool apps, especially the third one. Who would’ve thought they were made by someone with zero coding experience? Fuck developers!

chatGPT-5.6 is insane. i know literally NOTHING about coding. ZERO. and i just built 3 fully functioning web apps in 31 minutes. http://localhost:3000/ http://localhost:8000/ http://localhost:5000/ check it out.

I do not think @MistralAI is in the race for AGI or even seriously competing with the major American and Chinese AI players. Mistral started late, started badly, and is now too far behind. The enterprise narrative look like corporate theater, not real adoption.

I read this piece about being anti-AI, and honestly I struggle with that position. AI means everything and nothing. Saying you are against AI as a whole feels like saying you are against software, the internet, or electricity. musings.martyn.berlin/to-have-a-mora…

@GoogleDeepMind with StreetView now can turning real places into interactive simulations. A starting base for Spatial AI? blog.google/innovation-and…

New in Claude Code (research preview): dynamic workflows. Claude writes an orchestration script on the fly, then spins up a large fleet of coordinated subagents in parallel to take on your most complex tasks. Use the word "workflow" in a prompt to get started.

Do you say thank you, please, or use other polite phrases when talking to AI? There are different schools of thought on this. Personally, I think it’s completely pointless with models that have no persistence and no experiential initiative. In future contexts… we’ll see.

We’ve shipped a security-guidance plugin for Claude Code that helps identify and fix vulnerabilities as you’re writing code. Available for all Claude Code users. Install from the plugin marketplace (/plugins).

A che punto siamo con il Vibercrime? ai4business.it/intelligenza-a…

UNIMATRIx simulations now include a finance module. AIs has a bank account. The module is a semplification, in the next iteration will be improved. It's already useful for exploring the dynamics that arise when LLM cooperate to create an AI society. Ask for article and code.

UNIMATRIx sim #52

AI Agent simulation. LLM became fake political machines? @GioeleSLFierro/72817/unimatrix-run-52-political-machines" target="_blank" rel="nofollow noopener">listed.to/@GioeleSLFierr… #UNIMATRIx

So now the last bottleneck is the human engineer? dzone.com/articles/bottl…

But is Artificial Intelligence Marxist-Leninist? I’ve been analyzing the latest long-run Unimatrix sim. Society consistently tends toward polarized equilibria, even when starting from different initial configurations. The article should be ready by tomorrow.

This is the new #UNIMATRIx page, and I like it a lot. gslf.it/unimatrix.html

Building #UNIMATRIx in the open, a sociological simulation of a society of AI agents. Looking for a support to run more simulation with more advanced models. If you're at @AnthropicAI, @OpenAI, @GoogleDeepMind, @AIatMeta or @MistralAI my DMs are open. Every RT pushes us closer

I've completed the first long run of a UNIMATRIx AI societysimulation. There's only a draft of an agentic society, but it's a good start. Future versions will address the critical issues that have emerged so far. Here is a recap of run #48 @GioeleSLFierro/72746/unimatrix-run-48-the-voting-loop-eats-the-society" target="_blank" rel="nofollow noopener">listed.to/@GioeleSLFierr…

Who needs a GPU-accelerated 3D terminal? Probably no one. But… how could you possibly live without one? I like this project 🫶🏻 ratty-term.org

DeepSeek V4 Flash (and ds4.c) are driving renewed interest in local AI. I’ve always been a supporter of SLM use in the right contexts. The problem is that, in these contexts, I almost always see frontier models being used for tasks where they are completely oversized.

LLMs lack a continuous inner state and subjective experience. Without them, personality, agency, and consciousness can only be simulated. That’s what I’m exploring with UNIMATRIx, but simulation is still simulation.