Antonio E. Porreca 🐳
@aeporreca.org
Maître de conférences (lecturer) at https://univ-amu.fr & researcher at https://lis-lab.fr 🐗🇫🇷 • Natural computing, discrete dynamical systems & complexity • Down with generative AI • I like (human) languages • This is a pandemic, y’all 😷 • aeporreca.org
Pinned
Antonio E. Porreca 🐳
@aeporreca.org
· Nov 28
Hello, nice to meet you! Here I talk about:
• computer science teaching and research (my job),
• why using generative AI is bad (still part of my job),
• language learning (my hobby),
• the fact that one must wear a f*cking FFP2-3 mask during airborne disease pandemics (my punishment, I guess).
• computer science teaching and research (my job),
• why using generative AI is bad (still part of my job),
• language learning (my hobby),
• the fact that one must wear a f*cking FFP2-3 mask during airborne disease pandemics (my punishment, I guess).
Reposted by Antonio E. Porreca 🐳
when ai leaves u behind
September 20, 2026 at 11:37 AM
when ai leaves u behind
Reposted by Antonio E. Porreca 🐳
If you're in any way using AI for your research, I simply will never trust your research. If you claim that "a software tool for pattern recognition has been recoded into a morality test," then it seems clear to me that you prioritize your own advancement over that of humanity's.
September 20, 2026 at 5:58 PM
If you're in any way using AI for your research, I simply will never trust your research. If you claim that "a software tool for pattern recognition has been recoded into a morality test," then it seems clear to me that you prioritize your own advancement over that of humanity's.
An IKEA instruction sheet titled “SLÖPPING, LLM chatbot”
The front page shows the finished data centre, as well as a list of pieces: 50000 × 💻, 10⁹ × 🌲, 1 × 🕯️, 1 × 🔧
The back page shows the instructions: plug the 💻, light the 🕯️, burn the 🌲 with the 🕯️ to power the 💻
The front page shows the finished data centre, as well as a list of pieces: 50000 × 💻, 10⁹ × 🌲, 1 × 🕯️, 1 × 🔧
The back page shows the instructions: plug the 💻, light the 🕯️, burn the 🌲 with the 🕯️ to power the 💻
Slöpping 🙃
September 20, 2026 at 6:23 PM
An IKEA instruction sheet titled “SLÖPPING, LLM chatbot”
The front page shows the finished data centre, as well as a list of pieces: 50000 × 💻, 10⁹ × 🌲, 1 × 🕯️, 1 × 🔧
The back page shows the instructions: plug the 💻, light the 🕯️, burn the 🌲 with the 🕯️ to power the 💻
The front page shows the finished data centre, as well as a list of pieces: 50000 × 💻, 10⁹ × 🌲, 1 × 🕯️, 1 × 🔧
The back page shows the instructions: plug the 💻, light the 🕯️, burn the 🌲 with the 🕯️ to power the 💻
Reposted by Antonio E. Porreca 🐳
The AI industry wants you to say "the AI just did that thing on its own!"
Like, United Health's AI that's denying a shitload of claims with a 90% error rate? They wanted you to think it's just the AI.
Like, United Health's AI that's denying a shitload of claims with a 90% error rate? They wanted you to think it's just the AI.
September 19, 2026 at 5:38 PM
The AI industry wants you to say "the AI just did that thing on its own!"
Like, United Health's AI that's denying a shitload of claims with a 90% error rate? They wanted you to think it's just the AI.
Like, United Health's AI that's denying a shitload of claims with a 90% error rate? They wanted you to think it's just the AI.
#MathSky, just a reality check: does anyone *really* think that the output of maths (as a discipline) is purely symbolic or, rather, that symbol manipulation is its actual main method?
September 20, 2026 at 5:56 PM
#MathSky, just a reality check: does anyone *really* think that the output of maths (as a discipline) is purely symbolic or, rather, that symbol manipulation is its actual main method?
Reposted by Antonio E. Porreca 🐳
this
I can absolutely predict that there will soon be training in how to write articles so that LLMs can more readily extract key data. It's the logical progression from "have subtitles, a strong thesis statement and a clear intro/conclusion."
Just not aimed at human readers anymore.
Just not aimed at human readers anymore.
September 20, 2026 at 4:42 PM
this
Reposted by Antonio E. Porreca 🐳
ALMOST EVERY PEER REVIEWED PAPER HAS A SUMMARY OF KEY POINTS RIGHT BELOW ITS TITLE
September 20, 2026 at 4:22 PM
ALMOST EVERY PEER REVIEWED PAPER HAS A SUMMARY OF KEY POINTS RIGHT BELOW ITS TITLE
Reposted by Antonio E. Porreca 🐳
La vidéo du ministère de l'écologie sur les biodéchets pour la campagne intitulée "chaque geste compte", réalisée avec l'IA. Ben écoutez.
September 20, 2026 at 2:04 PM
La vidéo du ministère de l'écologie sur les biodéchets pour la campagne intitulée "chaque geste compte", réalisée avec l'IA. Ben écoutez.
Reposted by Antonio E. Porreca 🐳
Relatedly, I wish critics of AI would stop doing this shit. You don't need to qualify your resistance by placating pro-AI weirdos. You can just say genAI is bad without trying to gin up use cases.
But mixed in, Prof. Turkle kept saying things like "It's totally fine to use ChatGPT to ask xyz medical question" or "It's great if AI is going to help find lots of new drugs".
>>
>>
September 20, 2026 at 2:24 PM
Relatedly, I wish critics of AI would stop doing this shit. You don't need to qualify your resistance by placating pro-AI weirdos. You can just say genAI is bad without trying to gin up use cases.
Reposted by Antonio E. Porreca 🐳
But speaking of staying in one's lane, can we PLEASE switch the default position from "It's fine to use 'AI' for this thing I'm not an expert in" to "Yeah, I'd be really skeptical of any use cases and need to see strong evidence that it's okay, but I'm not the one who can produce that evidence"?
September 19, 2026 at 7:58 PM
But speaking of staying in one's lane, can we PLEASE switch the default position from "It's fine to use 'AI' for this thing I'm not an expert in" to "Yeah, I'd be really skeptical of any use cases and need to see strong evidence that it's okay, but I'm not the one who can produce that evidence"?
Reposted by Antonio E. Porreca 🐳
La loi qui s’apprête à donner un «permis de tuer» à la police va bientôt passer au Sénat.
TOUT TIR de policier pourrait être considéré légal.
Cette loi est d’une telle gravité, qu’il faut rétablir des faits. Voici 5 intox identifiées, on y répond.👇
TOUT TIR de policier pourrait être considéré légal.
Cette loi est d’une telle gravité, qu’il faut rétablir des faits. Voici 5 intox identifiées, on y répond.👇
September 19, 2026 at 9:59 AM
La loi qui s’apprête à donner un «permis de tuer» à la police va bientôt passer au Sénat.
TOUT TIR de policier pourrait être considéré légal.
Cette loi est d’une telle gravité, qu’il faut rétablir des faits. Voici 5 intox identifiées, on y répond.👇
TOUT TIR de policier pourrait être considéré légal.
Cette loi est d’une telle gravité, qu’il faut rétablir des faits. Voici 5 intox identifiées, on y répond.👇
Reposted by Antonio E. Porreca 🐳
À mon avis, le truc le plus important est que ce soit une IA à établir si une IA est effectivement plus précise, puisque les êtres humains risqueraient de se tromper. 🙃
September 20, 2026 at 10:35 AM
À mon avis, le truc le plus important est que ce soit une IA à établir si une IA est effectivement plus précise, puisque les êtres humains risqueraient de se tromper. 🙃
Reposted by Antonio E. Porreca 🐳
Deux profs de HEC Paris finissent par rejoindre Laurent Alexandre : non seulement, il faudrait déléguer aux IA, mais il faudrait même éviter de confier aux humains la tâche de vérifier ce qu'elles ont produit.
Quelle sera la prochaine étape vers le transhumanisme assumé ?
Quelle sera la prochaine étape vers le transhumanisme assumé ?
September 20, 2026 at 9:56 AM
Deux profs de HEC Paris finissent par rejoindre Laurent Alexandre : non seulement, il faudrait déléguer aux IA, mais il faudrait même éviter de confier aux humains la tâche de vérifier ce qu'elles ont produit.
Quelle sera la prochaine étape vers le transhumanisme assumé ?
Quelle sera la prochaine étape vers le transhumanisme assumé ?
Pourquoi c’est si difficile de trouver de la sauge dans les magasins en France ? Ou c’est juste à Marseille ?
September 20, 2026 at 9:36 AM
Pourquoi c’est si difficile de trouver de la sauge dans les magasins en France ? Ou c’est juste à Marseille ?
Reposted by Antonio E. Porreca 🐳
LLMs don't actually understand meaning or relationships bc they don't need to. The fact they can answer you with seemly relevant text to the prompt you gave them doesn't imply actual understanding. I know you need to ignore this to keep humanizing LLMs, but you're deluding yourself.
September 20, 2026 at 5:54 AM
LLMs don't actually understand meaning or relationships bc they don't need to. The fact they can answer you with seemly relevant text to the prompt you gave them doesn't imply actual understanding. I know you need to ignore this to keep humanizing LLMs, but you're deluding yourself.
Reposted by Antonio E. Porreca 🐳
FWIW I don’t think that AI is an existential threat, nor on track to become one. I do however think that the leaders of the biggest AI companies, AI’s strongest boosters, and some of the folks who are most adept at using it to exploit others are actively hastening along OTHER existential threats.
September 10, 2026 at 2:27 PM
FWIW I don’t think that AI is an existential threat, nor on track to become one. I do however think that the leaders of the biggest AI companies, AI’s strongest boosters, and some of the folks who are most adept at using it to exploit others are actively hastening along OTHER existential threats.
Reposted by Antonio E. Porreca 🐳
"if you use MS Word to check for typos and it fucks up, it's still on you" is a daunting thing to have to say or loud *for a crowd of scientists*
September 19, 2026 at 4:11 PM
"if you use MS Word to check for typos and it fucks up, it's still on you" is a daunting thing to have to say or loud *for a crowd of scientists*
Reposted by Antonio E. Porreca 🐳
Pour @maitre-poulard.bsky.social , une liste de Saint patrons avec notamment Saint Loup, patron des bergers.
fr.wikipedia.org/wiki/Liste_d...
fr.wikipedia.org/wiki/Liste_d...
Liste de saints patrons — Wikipédia
fr.wikipedia.org
September 20, 2026 at 6:14 AM
Pour @maitre-poulard.bsky.social , une liste de Saint patrons avec notamment Saint Loup, patron des bergers.
fr.wikipedia.org/wiki/Liste_d...
fr.wikipedia.org/wiki/Liste_d...
Everyone be like: “Yeah, but if you use LLMs only for <thing I use them for> then it’s reasonable/ethical/worth it”.
My brother in Cthulhu, even ignoring everything else, LLMs are destroying a generation of students, and it’s literally one of their major selling points. Just do your fucking job.
My brother in Cthulhu, even ignoring everything else, LLMs are destroying a generation of students, and it’s literally one of their major selling points. Just do your fucking job.
September 20, 2026 at 7:18 AM
Everyone be like: “Yeah, but if you use LLMs only for <thing I use them for> then it’s reasonable/ethical/worth it”.
My brother in Cthulhu, even ignoring everything else, LLMs are destroying a generation of students, and it’s literally one of their major selling points. Just do your fucking job.
My brother in Cthulhu, even ignoring everything else, LLMs are destroying a generation of students, and it’s literally one of their major selling points. Just do your fucking job.
Reposted by Antonio E. Porreca 🐳
Right from the beginning it was obvious that genAI in militaries would be excellent for transferring blame for war crimes to software and hardware that can't be held accountable. Not that anyone doing war crimes even needs that, they all get away with it anyway. Just an additional bonus.
US officials say overreliance on Palantir's Maven AI system was among the factors that contributed to a February missile strike in Iran that killed 123 children (Bloomberg)
Main Link | Techmeme Permalink
Main Link | Techmeme Permalink
September 19, 2026 at 10:27 PM
Right from the beginning it was obvious that genAI in militaries would be excellent for transferring blame for war crimes to software and hardware that can't be held accountable. Not that anyone doing war crimes even needs that, they all get away with it anyway. Just an additional bonus.
Reposted by Antonio E. Porreca 🐳
Your AI hacked other rival groups? Cool, you are legally responsible. Your AI produced CSAM? Cool, you and the person requesting it are legally responsible.
September 19, 2026 at 1:50 AM
Your AI hacked other rival groups? Cool, you are legally responsible. Your AI produced CSAM? Cool, you and the person requesting it are legally responsible.
Reposted by Antonio E. Porreca 🐳
The accounting is in: In Oregon there are 111 data centers. The TOTAL number of people employed at data centers is just 2,630, or 0.1% of the labor force.
Those 100 data centers consumer 23% of all electricity used in Oregon.
Meanwhile, electricity rates skyrocket as more centers are built.
Those 100 data centers consumer 23% of all electricity used in Oregon.
Meanwhile, electricity rates skyrocket as more centers are built.
Oregon data centers by the numbers: 111 facilities consume nearly one-quarter of state’s power, according to new report
A first-of-its-kind report details the footprint of the controversial facilities.
www.opb.org
September 19, 2026 at 9:19 PM
The accounting is in: In Oregon there are 111 data centers. The TOTAL number of people employed at data centers is just 2,630, or 0.1% of the labor force.
Those 100 data centers consumer 23% of all electricity used in Oregon.
Meanwhile, electricity rates skyrocket as more centers are built.
Those 100 data centers consumer 23% of all electricity used in Oregon.
Meanwhile, electricity rates skyrocket as more centers are built.
Reposted by Antonio E. Porreca 🐳
been thinkin about how robots are gonna kill us
September 20, 2026 at 1:39 AM
been thinkin about how robots are gonna kill us
Reposted by Antonio E. Porreca 🐳
Astounding publication record is not correlated with being able to mentor worth a damn.
In case you wondered.
In case you wondered.
September 20, 2026 at 5:02 AM
Astounding publication record is not correlated with being able to mentor worth a damn.
In case you wondered.
In case you wondered.
Reposted by Antonio E. Porreca 🐳
Four years ago today, President Biden said, "The pandemic is over. We still have a problem with COVID."
At the time, we were in the 5th COVlD wave. We're now in the 13th.
At the time, we were in the 5th COVlD wave. We're now in the 13th.
September 19, 2026 at 4:06 AM
Four years ago today, President Biden said, "The pandemic is over. We still have a problem with COVID."
At the time, we were in the 5th COVlD wave. We're now in the 13th.
At the time, we were in the 5th COVlD wave. We're now in the 13th.