Archives for February, 2025
The United States of Ruzzia + links
America first. Ruzzia firster. I'm afraid that, in the long run, China will eventually become the firstest. Note that in the current situation, Trump is merely the bull in a…
Quickest quickie: Claude 3.7 is here
Just before taking an 8-day pause from blogging, I noticed that Claude got a version upgrade. The official announcement, Claude Sonnet and Claude Code, is informative enough, but there are…
Me no know much, but running LLMs locally was disappointing
I never thought I’d be doing that, especially as it doesn’t make sense on a €400 laptop that lacks a proper GPU and VRAM: i3-1215U with Intel graphics (ADL GT2)…
Führer’s Grok 3 is promising―and expensive if you need more of it 🥴
It's a disgrace to have an LLM and an AI chatbot belonging to a complete asshole and a Nazi. But the AI model per se seems promising. What I mean…
It’s the automotive industry that might kill the BRD ➕ wahl.chat
If you were sleeping under a rock, Politico has it for you: German automakers have also bet big on the electrification of cars, investing billions in developing new models and…
The Amazon Kindle and stupid people: a storm in a teacup
There is one small change that will happen in relation to Kindle e-books, as noticed recently by people purchasing a book who wish to download it as a file and…
🤖 Mistral is a bad French “citizen”
I had a mental blank and could not recall the name of a French bookstore from which, many years ago, I purchased an e-book. It started with M, but what…
How does a LLM work? These videos won’t help that much ➕ Microsoft hallucinates!
All the YouTube videos about how AI works, with stress on the chatbots based on LLMs, are superficial. It can’t be otherwise for two reasons: any such presentation is popularized…
SEO: Is Claude on my side, or it’s parroting me?
There is one thing that I could never understand. I asked ChatGPT about it on March 17, 2023, and I wasn't satisfied with the results. Now I just asked Claude,…
You Don’t Fool Me
I used to be annoyed by the huge influx of hoaxes, fake news, and especially fake quotations long before any LLM was born. And today's chatbots can hallucinate quotations for…
The Führer’s minions have restored my account
While people and organizations are leaving X, I just got restored. It’s the end of the world. (It wasn’t automatic; I contested it again, and in 2 minutes I was…
I tried to understand “no contest” without consulting Wikipedia
I’m sick of chatbots because I'm using them since March 13, 2023. More recently, I thought DeepSeek to be OK, then horrendous, and finally unable to admit guilt. Thankfully, I…
DistroWatch Weekly strikes again!
It is happening again. Why do I even bother? I can't stand idiots, and I run into them. OK, so I knew that DWW is censoring me (after the facts,…
Et tu, Claude? Having fun with AI
Notwithstanding all my other previous posts on AI, while playing with my short list of AI chatbots, starting with Claude, I found that sometimes Claude can be completely off-track. I…
Swimming upstream against the AI current—updated!
I had to give the seal of approval precisely to the only one of the 3 AI agents I currently use that cannot search the web! Why? Well, because searching…
Was I wrong in my take on packages vs. Flatpaks & snaps? Updated!
I don’t even remember how many times I insisted—because a Luddite has to be conservative, and Linux used to observe the KISS principle anyway—that the right thing to install software…
🤖 How to properly use AI as a software developer: through its API
Why, indeed, you shouldn’t use an AI chatbot to write non-open-source software, bar some code snippets and other minor adjustments, but people do it—with or without their employer's knowledge. It’s…
🤖 The biggest European IA conference nobody cares about
I’m not entirely sure whether what the title says is what I believe, but the fact is that El Presidente Macron El Primero wants to make France a big name…
🤖 A curt explanation on my latest choice of AI systems
Few people managed to read my two previous mega-posts on AI models (from Feb. 1 and Feb. 6), with added comments and updates to the text. They were supposed to…
🤖 AI: from LLMs to Enslavement ● ChatGPT lies about its Search!
When I wrote my previous post about AI, triggered by DeepSeek’s success, I did touch most aspects that I considered relevant for the current AI bubble, centered on LLMs and…