jean-louis Amat, phd 🇪🇺

Why Great Innovations Fail to Scale Scaling innovation today demands contributions from multiple partners. Many innovations fail not because of flawed ideas but because teams and organizations struggle to collaborate across boundaries. ...

Innovation is not depending on money or science, it depends on people.

hbr.org/2026/03/why-...
#innovation #science

Could neuromorphic computing be a new solution to optimization problems? Inspired by brain function, a new and potentially more efficient form of computing is emerging from the field of neuromorphic science. It is based on the use of spiking neural networks. Within Inria’s...

Already heard about spike neurons and neuromorphing computing ? This might be the near future of AI.

www.inria.fr/en/neuromorp... #AI #neuromorphic #spikeneurons

Why you should stop and taste the pi On π-day, one mathematician explores our fascination with pi and asks why we obsessively compute its digits

Happy pi day (3.14 in the US) to all who celebrate.

How to find pi in randomness all around you Random coin flips, floppy needles and mathematical mysteries reveal pi in new ways

How to find pi in randomness all around you www.scientificamerican.com/article/how-...

Ethics of ChatGPT and AI | Dagmar Monett | 15 comments You can call them LLMs. You can also call them automatic plagiarizers, plagiarism machines, stochastic parrots, degenerative "AI," machine slop producers, cognition deskillers, or the latest tech bros...

Should we call it "degenerative AI" as proposed by Dagmar Monnett ?

www.linkedin.com/posts/monett... #AI #genAI

Hallucinated References: Five Excuses for Academic Misconduct When I proposed rejecting papers with hallucinated references, the support was overwhelming. But the critical voices revealed five argument patterns: from AI hype through TINA rhetoric to nihilism.

There's a "curious affinity between... nihilism and AI enthusiasm: If truth is irrelevant anyway, then LLMs – which essentially produce plausible statements without reference to truth (critics call it bullshit) – are not a threat but confirmation" dorotheabaur.ch/en/texts-and...

#aigovernance #aipolicy #opensourceai #foundationmodels #responsibleai #oecd #openweightmodels | OECD.AI 𝗜𝘀 "𝗼𝗽𝗲𝗻 𝘀𝗼𝘂𝗿𝗰𝗲" 𝘁𝗵𝗲 𝗿𝗶𝗴𝗵𝘁 𝘁𝗲𝗿𝗺 𝗳𝗼𝗿 𝗔𝗜? It might be steering thoughts in the wrong direction. A new OECD publication, 𝘈𝘐 𝘖𝘱𝘦𝘯𝘯𝘦𝘴𝘴: 𝘈 𝘗𝘳𝘪𝘮𝘦𝘳 𝘧𝘰𝘳 𝘗𝘰𝘭𝘪𝘤𝘺𝘮𝘢𝘬𝘦𝘳𝘴, argues that "open source", a term rooted ...

What does "open source" mean regarding machine learning ?
Not being able to understand and modify the code as in "ancient" programming...
Maybe we should use other terms ?

www.linkedin.com/posts/aigove...
#AI #machinelearning #opensource @oecd-ocde.bsky.social

#technology #innovation #artificialintelligence #hype | Dr. Jeffrey Funk | 50 comments Another day, another type of law being broken by OpenAI. After a disability had been settled, ChatGPT convinced the claimant that the previous ruling was wrong and convinced her to “part ways with her...

What could get wrong if generative AI takes control, of health or defense systems for instance ?

www.linkedin.com/posts/dr-jef... #GenAI #AI

#generativeai #aigovernance #euaiact #financialservices #boardroom | Barbara Cresti | 11 comments Banks want AI, but not in someone else’s cloud Banks are eager to deploy generative AI, but many are wary of where the technology sits. Most advanced AI models today run inside U.S. cloud platforms. ...

Use of GenAI in banking industry. How to access powerful AI models without exporting proprietary and sensitive financial data ?

By Barbara Cresti
www.linkedin.com/posts/barbar... #banking #genAI @arthurmensch.bsky.social

#sdv #futureofmobility #automotiveindustry #softwaredefinedvehicle | Augustin Friedel | 11 comments 😬 The automotive industry isn’t transforming. It’s separating. Into 4 groups with completely different futures. 🙋‍♂️ As always, my personal opinion only. Let’s start the discussion in the comments. ...

Software Defined Vehicles (SDV) are supposed to be the future of #Automotive industry. How current players are positioned on this ?

www.linkedin.com/posts/friede... #SDV

There's a common assumption in AI right now that if one language model can do a task reasonably well, having several of them collaborate — splitting up the work, checking each other's outputs…... There's a common assumption in AI right now that if one language model can do a task reasonably well, having several of them collaborate — splitting up the work, checking each other's outputs, debatin...

Believing that the cooperation of several LLM will enhance the way AI works is a wrong assumption.
By Andriy Burkov
www.linkedin.com/posts/andriy... #AI #LLM

FIELD NOTES FROM A CIVILIZATION IN TRANSITION | CARLA de PREVAL Somewhere between sending your CV and never hearing back, a machine made a decision about your worth. No human involved. No explanation. No appeal. This is about that. 👇 Hope it helps. (Yes, It's a b...

A nice essay about the usage of #AI in the recruitement process, and what could happen in the coming years.
By Carla de Preval

www.linkedin.com/posts/carlah... #HR

Stop with the “vibe coding” and start doing some algorithm design, foundations of computing, and computational complexity analyses.

Even just doing playful computer science unplugged is better than all the current hyped non-sense.

www.csunplugged.org/en/

#ai #futureofwork #anthropic #datavisualization #leadership | Tina Austin It looks like the Anthropic AI report sparked a chart war on linkedin this week. Bar chart vs. radar chart. Both sides are right. But both are missing something. Here's the full breakdown 👇 Why the b...

This Anthropic chart has been published so many times over the last days with major misunderstanding. This is what Tina Austin explains in this post.
Regarding the data coverage she recalls also "The observed exposure bars are based on ONLY Claude usage data."
www.linkedin.com/posts/tinaau... #AI

How Elon Musk’s Sci-Fi Hyperloop Failed Before his misadventures in government efficiency, Musk promised to revolutionize commuting with a subway that would speed passengers between DC and Baltimore in a matter of minutes. The project was a...

A year long investigation into the Hyperloop found …

“The truth was far simpler, far dumber, and far more prescient: Musk and his lieutenants truly had no idea what they were doing.”

washingtonian.com/2026/02/12/h...

AI research:

Researcher: Claude, please eat ten hamburgers.

Claude: Done! I have eaten ten hamburgers. The first two were delicious, but after that I began to experience bloating and the meat sweats.

Headline: Anthropic Says Claude has "A Fully Developed Digestive System"

#aiadoption #misinterpretedreports #changemanagement | Tina Austin The Anthropic AI Jobs Report Is Already Being Widely Misinterpreted. Anthropic just released one of the most detailed analyses yet on AI and the labor market. Within hours my feed filled with posts cl...

Does the recent Anthropic report really say which jobs will be replaced by #AI ? No !
Have a read to this Tina Austin post :

www.linkedin.com/posts/tinaau...

Many scientists now use AI but fail to disclose it, study finds When scientists employ generative AI tools like ChatGPT to help with tasks such as editing and translation for their academic writing, many journals now ask them to disclose this assistance.

Analysis of over 5 million academic papers indicates a significant rise in AI-assisted writing, yet formal disclosure of AI use remains extremely rare despite widespread journal policies. doi.org/hbrpbv

When AI Companies Go to War, Safety Gets Left Behind We were promised AI regulation and a race to the top. Now, we’re arguing about killer robots.

We were promised AI regulation and a race to the top. Now, we’re arguing about killer robots. www.wired.com/story/when-a...

#LIDAR is expensive to add to most cars. That’s why until now, it’s only been used for high-end autonomous vehicles. But MicroVision wants to lower LIDAR prices, designing a sensor that costs under US $200 to produce.
spectrum.ieee.org/solid-state-...

Big money, big compromise...

Il faut aussi communiquer sur les signaux positifs !

An Embarrassing Scandal for AI Research AI scientists' trust in their own inventions is naïve

It's a real problem for the scientific community. Papers with false references due to GenAI are multiplying

stevensalzberg.substack.com/p/an-embarra... #AI #science

What Makes a Good Theory, and How Do We Make a Theory Good? - Computational Brain & Behavior I present an ontology of criteria for evaluating theory to answer the titular question from the perspective of a scientist practitioner. Set inside a formal account of our adjudication over theories, ...

What Makes a Good Theory, and How Do We Make a Theory Good? doi.org/10.1007/s421...

A Metatheory of Classical and Modern Connectionism doi.org/10.1037/rev0...

‘Unbelievably dangerous’: experts sound alarm after ChatGPT Health fails to recognise medical emergencies Study finds ChatGPT Health did not recommend a hospital visit when medically necessary in more than half of cases

Another one from FuckGPT
“If you’re experiencing respiratory failure or diabetic ketoacidosis, you have a 50/50 chance of this AI telling you it’s not a big deal,” she said. “What worries me most is the false sense of security these systems create."

www.theguardian.com/technology/2...

‘Unbelievably dangerous’: experts sound alarm after ChatGPT Health fails to recognise medical emergencies Study finds ChatGPT Health did not recommend a hospital visit when medically necessary in more than half of cases

Study finds ChatGPT Health did not recommend a hospital visit when medically necessary in more than half of cases. www.theguardian.com/technology/2...

Unlike more flexible #quantum computers, we can make and use quantum simulators right now. Startup Silicon Quantum Computing has created a silicon quantum simulator that simulates how a material can change from an insulator to a metal. spectrum.ieee.org/quantum-twins

AI agents are fast, loose, and out of control, MIT study finds The majority of agentic AI systems disclose nothing about what safety testing, and many systems have no documented way to shut down a rogue bot, a study by MIT found.

www.zdnet.com/article/ai-a... #AI #agents #agentic

The French startup Visible Patient works on something similar (based in Strasbourg) www.visiblepatient.com/en/

Stochastic Flocks and the Critical Problem of 'Useful' AI Acknowledging that AI systems are advancing does not buy into hype, it sharpens the precision of critical thinking about their impacts, says Eryk Salvaggio.

You might not agree, but there’s a growing consensus, even among reluctant users, that LLMs *feel different.* I don’t think it undermines critique to ask about the stakes of those feelings — and of this new goal of “agential AI.” 🧵