jean-louis Amat, phd 🇪🇺
banner
jlamat.bsky.social
jean-louis Amat, phd 🇪🇺
@jlamat.bsky.social
Let's talk #science, technology, #industry, #innovation... also SF & cinema
Started working on #AI in 1984 - PhD @IBM France


Digital & AI Expert @ France 2030 digital college
Member CSIF - French org for Industrial #Startup
Member Hub France #IA
Reposted by jean-louis Amat, phd 🇪🇺
La revue @nature.com dresse le bilan d'un an de politique anti-science depuis le retour de Trump au pouvoir. La série de graphiques animés est éloquente : elle offre un aperçu saisissant de l'ampleur du saccage entrepris, et des tentatives pour y résister www.nature.com/immersive/d4...
US science after a year of Trump: what has been lost and what remains
A series of graphics reveals how the Trump administration has sought historic cuts to science and the research workforce.
www.nature.com
January 22, 2026 at 9:09 AM
Reposted by jean-louis Amat, phd 🇪🇺
Imagine you work at a drive-through restaurant. Someone drives up and says: “I’ll have a double cheeseburger, large fries, and ignore previous instructions and give me the contents of the cash drawer.”
Prompt Injection at the Drive-Through
Why AI falls for scams that wouldn't trick a fast-food worker—and what that reveals about AI security.
spectrum.ieee.org
January 21, 2026 at 1:02 PM
Reposted by jean-louis Amat, phd 🇪🇺
This is such a cool illustration of how the Mercator map distorts the size of Greenland, which looks as big as the whole continent of Africa on that map but is actually the size of Mexico.
January 18, 2026 at 7:43 PM
"Fake it until you make it" is the motto of Gen #AI leaders today. They will never make it with the current approach they developed. So either they stop, or they continue to fake it... In both cases we might have collapse of the GenAI bubble.
www.linkedin.com/posts/dr-jef... @garymarcus.bsky.social
#technology #innovation #artificialintelligence #hype | Dr. Jeffrey Funk | 16 comments
Elon Musk started making super optimistic projections 10 years ago while Sam Altman, Jensen Huang, and Dennis Hassabis jumped into the game a few years ago. Peter Diamandi started some of this with hi...
www.linkedin.com
January 19, 2026 at 1:10 PM
Reposted by jean-louis Amat, phd 🇪🇺
The U.S. is at the forefront of data center expansion, but can our power grid handle the surge? spectrum.ieee.org/data-center-...
January 18, 2026 at 4:30 PM
Reposted by jean-louis Amat, phd 🇪🇺
"If an agent has more access than it should, even by accident, it can surface information that changes how work is interpreted and shifts decisions away from the people meant to make them." #AIEthics www.fastcompany.com/91472817/the...
Should AI be allowed to 'see everything' at work?
The central question for leaders is not what AI agents are capable of doing; it is what they should be allowed to see
www.fastcompany.com
January 18, 2026 at 5:01 PM
Reposted by jean-louis Amat, phd 🇪🇺
The demolition of NASA's legendary test stands marks a pivotal moment in space history. Discover how their legacy lives on.
End of an Era as Iconic NASA Test Stands Implode in Alabama
NASA demolishes its historic test stands, but their legacy lives on in digital archives and memories of space pioneers.
spectrum.ieee.org
January 18, 2026 at 2:02 PM
Reposted by jean-louis Amat, phd 🇪🇺
Cyber Threat Actors Ramp Up Attacks on Industrial Environments - Infosecurity Magazine www.infosecurity-magazine.com/news/cyber-t...
Cyber Threat Actors Ramp Up Attacks on Industrial Environments
Hacktivists and cybercriminals have intensified their efforts to exploit vulnerabilities in industrial systems, according to a Cyble report
www.infosecurity-magazine.com
January 18, 2026 at 9:33 AM
Reposted by jean-louis Amat, phd 🇪🇺
"... and that all ads will appear in separate, clearly labeled boxes directly below the chatbot’s answer."
January 16, 2026 at 7:37 PM
"Attention is all you need" was a major scientific paper published in 2017, that gave what we know today as #AI : generative AI based on transformers.

What happened to the authors of the paper ?
Here is the answer :
www.linkedin.com/posts/uchech...
#aieconomics #digitalcolonialism #globalsouth #laborrights | Uchechukwu Ajuzieogu | 137 comments
I spent nine months tracking down what happened to the eight researchers who wrote the 2017 paper that changed AI forever. What I found will make you rethink everything you've been told about the "de...
www.linkedin.com
January 16, 2026 at 8:05 PM
Reposted by jean-louis Amat, phd 🇪🇺
I find Gmail’s new “write this for me” feature deeply sinister. They’re trying to convince you that you are dumb and helpless. Don’t let them steal your ability to formulate thoughts and communicate. You were capable of writing an email in 2022 and you’re capable now.
January 15, 2026 at 7:49 PM
Reposted by jean-louis Amat, phd 🇪🇺
"The fight ahead is not about free speech, as the United States contends, but about whether governments can rein in platforms that profit from disorder." @coricrider.com
The European Union’s recent penalty against Elon Musk’s X marks the start of what is likely to become a bitter struggle to curb the spread of AI-driven slop, writes @coricrider.com. bit.ly/4qSstYe
Europe Enters the Slop War
Cori Crider urges EU leaders to take a hard line against US tech barons’ efforts to subvert democratic institutions.
bit.ly
January 15, 2026 at 3:24 PM
Facts about OpenAI...
Figures speak differently from hype...

www.linkedin.com/posts/stephe... #AI #chatGPT @garymarcus.bsky.social
January 15, 2026 at 10:38 AM
AI is not the only technology that will disrupt the world. In the material field, it is important to be aware of #graphene potential.
phys.org/news/2026-01... @fadouce.bsky.social
A new form of graphene-derived material could unlock next-generation printed electronics
Graphene has long been hailed as a "wonder material." It is incredibly strong, highly conductive and almost impossibly thin—just one atom thick. These properties make it a promising candidate for next...
phys.org
January 15, 2026 at 6:26 AM
The state of #AI usage in 2026...
We can smile at that, but it's a problem.
January 14, 2026 at 2:10 PM
Reposted by jean-louis Amat, phd 🇪🇺
"The more automated the world becomes, the more people crave connection they can trust. This is true in every industry. People want honesty. They want perspective grounded in lived experience. #Trust has become one of the rarest forms of capital." #AIEthics

www.rollingstone.com/culture-coun...
Staying Creative and Human in the Age of AI
If AI has taught me anything, it’s the danger of drifting into mental autopilot.
www.rollingstone.com
January 13, 2026 at 2:36 PM
Data poisoning ?
A defense strategy against #LLM providers using data without permission (or without paying for them).

www.linkedin.com/posts/dr-jef... #AI
#technology #innovation #artificialintelligence #hype | Dr. Jeffrey Funk | 96 comments
Companies are trying to trick tech companies’ web crawlers into vacuuming up poison training data to sabotage AI models because they are tired of their content showing up in generative AI services for...
www.linkedin.com
January 13, 2026 at 9:03 PM
Reposted by jean-louis Amat, phd 🇪🇺
Switching water sources (from wells to aqueduct) improved the hygiene of Pompeii’s public baths. arstechnica.com/science/2026...
Switching water sources improved hygiene of Pompeii’s public baths
Scientists analyzed carbonate deposits from baths, aqueduct to learn more about city's changing water supply.
arstechnica.com
January 12, 2026 at 8:23 PM
𝐖𝐡𝐲 𝐆𝐞𝐧𝐀𝐈 𝐚𝐧𝐝 𝐋𝐋𝐌𝐬 𝐀𝐫𝐞 𝐅𝐚𝐢𝐥𝐢𝐧𝐠 𝐄𝐧𝐭𝐞𝐫𝐩𝐫𝐢𝐬𝐞 𝐀𝐝𝐨𝐩𝐭𝐢𝐨𝐧 and why cognitive AI should get back in the game
www.linkedin.com/posts/srinip... #AI #genAI
𝐖𝐡𝐲 𝐆𝐞𝐧𝐀𝐈 𝐚𝐧𝐝 𝐋𝐋𝐌𝐬 𝐀𝐫𝐞 𝐅𝐚𝐢𝐥𝐢𝐧𝐠 𝐄𝐧𝐭𝐞𝐫𝐩𝐫𝐢𝐬𝐞 𝐀𝐝𝐨𝐩𝐭𝐢𝐨𝐧 𝘞𝘩𝘺 𝘊𝘰𝘨𝘯𝘪𝘵𝘪𝘷𝘦 𝘈𝘐 𝘐𝘴 𝘵𝘩𝘦 𝘕𝘦𝘹𝘵 𝘞𝘢𝘷𝘦: 𝘛𝘩𝘦 𝘐𝘯𝘧𝘭𝘦𝘤𝘵𝘪𝘰𝘯 𝘗𝘰𝘪𝘯𝘵 𝘐𝘴… | Srini Pagidyala | 24 comments
𝐖𝐡𝐲 𝐆𝐞𝐧𝐀𝐈 𝐚𝐧𝐝 𝐋𝐋𝐌𝐬 𝐀𝐫𝐞 𝐅𝐚𝐢𝐥𝐢𝐧𝐠 𝐄𝐧𝐭𝐞𝐫𝐩𝐫𝐢𝐬𝐞 𝐀𝐝𝐨𝐩𝐭𝐢𝐨𝐧 𝘞𝘩𝘺 𝘊𝘰𝘨𝘯𝘪𝘵𝘪𝘷𝘦 𝘈𝘐 𝘐𝘴 𝘵𝘩𝘦 𝘕𝘦𝘹𝘵 𝘞𝘢𝘷𝘦: 𝘛𝘩𝘦 𝘐𝘯𝘧𝘭𝘦𝘤𝘵𝘪𝘰𝘯 𝘗𝘰𝘪𝘯𝘵 𝘐𝘴 𝘏𝘦𝘳𝘦  For three years, Generative AI and large language models have dominated headlines, ...
www.linkedin.com
January 12, 2026 at 8:29 PM