Профиль

@funnytechguy.bsky.social I mean, AFAIK, every US president since Reagan as been a terrorist. Some have even got condemned by UN and IC for that. Trump is an end product of decades of US policies based in obsessions, mostly about the commies coming for them moneys.

@swingingstorm.bsky.social If you like it there’s also in the same collection: Guatemala, Nicaragua, Salvador,Panama,.., and many, many more.

Dans ce pays on est tellement incapable de tracer l’argent public dont on abreuve les entreprises qu’on se laisse des fourchettes de 100 milliards d’€ en revanche on te demande de justifier que tu mérites chaque euro de tes 570€ de RSA

The questions we should ask are the ones having meaning for humanity regarding this overflow of AI in our daily lives. It is already ubiquitous, and causing harm in many ways we aren’t prepared for. So to extract ourselves from the hype, we should stop talking like it’s human, and care for humans.

So we shouldn’t be deceived by the outlandish statements of their creators - claiming AGI, whatever it means -, and we shouldn’t ask ourselves if it is indeed a « new form of consciousness », or wether we should protect it legally like one.

The abuse of anthropomorphism around AI isn’t only influencing people who have little knowledge on these technological subjects, it is also orienting the broader conversations and debates about it. AI is just a tool. But it’s pushed recklessly in every fields with a dangerous narrative.

And I mean, why not? These are said to have brain like structures, gives you answers like they’re a friend, and have the knowledge of every lawyers and doctors in memory. Why not taking their advice? The industry says that soon they will replace many of these professionals. Even on the battlefield.

They can mimic human language, art or psychology, and they can’t produce on their own without the human consciousness nudging them to. But still we can see a growing pattern of people taking LLMs for what they’re not.

These are still machines doing « computes », they have no moral compass, only guardrails, they have no empathy, even if they’re polite and inclined to validate, they are just multidimensional representation of a statical program looking for an answer in a very large database.

However the way you describe them, these tools are not humans or human like. They shouldn’t be described as such and even less treated as such. This lexical push from the tech Industry is in part a business move, and another part the result of a « Pygmalion syndrome » of their creators.

My real concern at this point is the analogies pushed by the AI industry. Wether it’s neuroscience, psychology, or even philosophy, it’s too often anthropomorphic.

They chose to push a language from an analogy, to give the sense of urgency, against another, rooted in a culture full of wordplay, puns and metaphors. On many levels I understand. It makes the approach easier and the overlap with intelligence and security is obvious. But, I digress.

The language we use is conditioning our way of thinking. I remember at the time, when Cybersecurity emerged, I hated that it chose to use the lexicon of the defense/intelligence industry over the hackers lingo, when able to choose from.

Pour ceux qui souhaitent s'opposer à l'exploitation de leurs données de santé par Doctolib, c'est avant fin juillet et c'est ici : www.doctolib.fr/privacy-sett... Pur un résumé des arguments des uns et des autres : www.franceinfo.fr/replay-radio...

There is a possibility that the statistical mashup for this particular piece has been partly made from the work of a struggling artist somewhere on the planet. Until we can’t prevent that, meaning having controls on the training datasets, we will refer to any artistic AI product as SLOP.

@4017xan.bsky.social I see 2 problems here personally. First, this must not be advertised as a photo. It is clearly AI. Second, if AI can be this tool that makes wonders, it is because it has stolen the work or so many during training.

🔴🗣️ « Ce soir, la Macronie va tenter de classer la pétition contre le permis de tuer XXL qui a déjà recueilli plus de 720 000 signatures. Nous ne les laisserons pas l'enterrer en catimini. ➡️ En direct à 21h : videos.assemblee-nationale.fr/direct.19384... » - @mathildepanot.bsky.social

@yourdigitalbrain.bsky.social I call BS. Someone had already painted the fake sky. Plus I’ve never seen videos of this Eratosthene guy you say existed with the dinosaurs. The truth is we all live in a Petri dish!

@zeenatdagrant.bsky.social You have the same with Sysinternals ‘handle.exe’. Simply look for the drive assigned letter. Like PowerToys you have to install first. But I think you can also get that handle with ‘resmon.exe’ (the Resource Monitor). Run resmon and check for the handle in CPU tab.

lsof is your friend.