Profile avatar
duperrin.com
People/Organization/Tech professional and enthusiast. Proud Parisian. Travel addict. Blogging and Blueskying in 🇫🇷and 🇬🇧 Duperrin.com Duperrin.com/english
63,470 posts 203 followers 87 following
Prolific Poster

L’IA révolutionne l’éducation, mais son impact dépend de notre usage. Pour éviter une dépendance passive, il est essentiel de développer esprit critique, curiosité et maîtrise des requêtes.

An open space is just a chaotic factory. Poor communication and lack of workload visibility turn everyone into a bottleneck—overloaded, ineffective, and at risk of harming both themselves and others. The good news? With a bit of discipline, you can regain control.

Déjà intégrée via des cas ciblés (facturation, logistique…), l’IA générative reste en phase d’expérimentation. Le passage à l’échelle se construit, avec des perspectives pour 2025. Comme pour l’IoT autrefois, difficile de mesurer aujourd’hui l’impact réel derrière les promesses.

Some tasks are made for robots but given to humans by default. Others derive value from being done by humans. AI promises to free workers from repetition but mostly erodes connection-driven work—critical in health, education, and law.

Digital now eats up 27% of the planet’s CO₂ budget, 18.7% of mineral resources, and 15.3% of fine particle emissions—and it’s only getting worse with AI. Could we change course? No. This industry is a pillar of global power. Cutting its impact? Unthinkable.

The real shift is happening with autonomous AI agents that plan, decide, and automate complex tasks. Multi-agent systems boost efficiency and are already used in finance, e-commerce, and HR. The key is to know how we integrate AI strategically to enhance competitiveness without disruption.

To boost productivity, companies reduce bureaucracy and hierarchy. But instead of mass layoffs, they should focus on talent density, accountability, and innovation. Drastic cuts don’t ensure success. Agile teams, a learning culture, and performance-driven management are the real keys.

Companies strive for operational excellence, yet few link it to soft skills, perhaps due to its industrial roots. But in a world where collaboration is a production system, remote work is the norm, and continuous improvement is essential, soft skills are no longer optional. bit.ly/3ETu25D

Documenter des cas d’usage figés pour une technologie en perpétuelle évolution ? Illusoire. En plus d’être vite obsolètes, ils n’engagent personne et renforcent le rejet. Plutôt que d’imposer des scénarios rigides, il faut donner aux équipes les clés pour expérimenter et s’approprier l’IA.

Is Generative AI a bubble? It sure looks like one. • Inflated user and revenue figures • Overblown projections vs. weak profits • No killer app, no real impact • A financial black hole The hype keeps investors in, but the crash will be brutal.

The Paris AI summit confirmed what was already clear: AI has become a geopolitical battleground. Like Berlin in 1949, it is not the subject but the stake in a clash between opposing blocs. It’s no longer about business—it’s about ideology, dominance, and power.

Is outsourcing cognitive tasks good? Like calculators reducing mental math, AI may lead to “cognitive outsourcing.” This frees mental bandwidth but risks weakening critical skills. AI can balance cognitive overload, but some abilities are too vital to lose—they’re humanity’s “core business.”

Are we doomed to inefficiency? Everything has changed in a century—work, production, tools—yet we still operate like a 1930s factory. The diagnosis and solutions have been clear for 50 years, but little action is taken. Frustrating, but I had to see it through.

Tech is rotting—and it’s profitable. Meta, Google, Microsoft… all make billions by making their products worse. Google breaks Search for ads, Facebook hides friends for engagement, Apple profits from scams. Tech media looks away, chasing trends. We deserve tech that works for us, not against us.

Traditional #HR is disappearing, replaced by a strategic trio: HR strategists, talent scientists & HR technologists. AI personalizes careers, training & coaching, while agile, skill-based organizations replace rigid hierarchies. The future isn’t just #AI , it’s AI plus better human #leadership.

Pour Luc Julia, co-créateur de Siri, l’IA générative n’est qu’une évolution des modèles existants, sans réelle créativité. Il dénonce l’exploitation des artistes, les risques du jailbreaking et des régulations inadéquates. Il prévoit la fin des LLM, remplacés par des IA spécialisées et hybrides.

Generative AI in the workplace delivers fewer benefits than expected. Adoption is slow, the “magic” effect fades, and gains are mostly seen in simple use cases, not complex processes. Shadow AI declines, but hope remains with the rise of agentic AI, which could unlock new potential.

Why are tech products called solutions when they often create new problems? From typewriters to AI, technology has been sold as a miracle fix, yet it only extends our capabilities. We anthropomorphize it, forgetting it merely reshapes social and power dynamics.

Meta abandonne le fact-checking, une décision critiquée comme une capitulation mais défendue au nom de la liberté d’expression. Un article qui plaide pour un équilibre entre liberté et responsabilité, inspiré du modèle de la presse.

Businesses must move from centralization to autonomy, where managers facilitate rather than command. This shift replaces authority with contribution, turning the company into an ecosystem that empowers employee skills Inspired by servant leadership, it becomes an “enterprise as a service” .

L’IA générative peine à prouver son ROI. Une étude de Wavestone et la French Tech Grand Paris montre une adoption lente, des défis technologiques et stratégiques, et un ROI difficile à mesurer.

Le sommet de Paris sur l’IA a montré une myopie : on parle d’investissements sans voir la bataille géopolitique. La tech n’est plus un marché, mais un terrain d’affrontement idéologique où les entreprises sont des pions. Cédric Durand y voit un « techno-féodalisme », renforcé par le duo Trump-Musk.

De moins en moins de jeunes veulent devenir managers. Selon une étude de Robert Walters, 52 % des moins de 30 ans refusent ces responsabilités, invoquant charge mentale et quête d'équilibre. www.rts.ch/info/societe...

Generative AI reduces cognitive effort but weakens critical thinking. The more we trust it, the less we question its outputs. Knowledge workers shift from creators to supervisors, risking cognitive laziness. AI should be treated as a hypothesis generator, with verification and critical review.

Paris AI Summit: Power Play or Progress? The real issue wasn’t AI but global dominance. Europe tries to catch up, but funds come from the US, UAE, and Canada. Ethics take a backseat as regulations weaken. Investments favor infra over innovation.

LinkedIn: A Lost Cause? Back in 2021, I already felt disenchanted. Now, I’m near the breaking point. The tool hasn’t evolved, but behavior has, and no one seems to care. It’s treated as a necessary evil. The real issue? LinkedIn does nothing to protect its essence a risk that may eventually backfire

Can individual aspirations and business needs truly align? In a chaotic work world, many seek a lifeline: meaning. But companies have goals, employees have their own ambitions—can they all move in the same direction? Perfect alignment is a myth, but narrowing the gap? That’s possible.

Pointer du doigt ceux qui prennent l’avion ou ne trient pas tout en prônant un usage immodéré du numérique, c’est de l’hypocrisie. Le digital émet déjà deux fois plus de CO₂ que l’aviation et engloutit 40 % de notre budget carbone. Une requête IA consomme 10 fois plus qu’une recherche Google.

European businesses have big AI ambitions, but is it realistic. Europe has assets, but without a real institutional push, it’s just wishful thinking. A tech Airbus? Still waiting. Will institutions step up, or will bureaucracy win again?

Businesses talk about sustainability, but digital’s environmental impact is often ignored. Digital usage now surpasses air travel in emissions, yet professional use is downplayed. Unlike factories or planes, it’s invisible but just as harmful. Technology won’t fix what it fuels—behavior must change.

Before tearing down a load-bearing wall, ask why it was built. The tech sector's obsession with disruption ignores the risks of dismantling systems without understanding their purpose. In critical sectors, reckless change can cause chaos.

“Streamlining” often means adding complexity rather than simplicity. While meant to enhance efficiency, it frequently leads to more bureaucracy and frustration. A phrase I’ve used often, it rarely delivers the intended benefits. Theory and reality don’t always align when processes are “optimized.”

Can AI succeed where digital transformation failed? I doubt it. Companies mistook digital transformation for digitization, reinforcing silos instead of breaking them. AI promises to unify data and collaboration, but silos are cultural and political, not just technical. Let’s check back in 10 years.

L'art d'enfoncer des portes ouvertes.... Les solutions de 1970 fonctionnaient en 2000 ? Les solution de 1940 fonctionnaient en 1970 ? Et moi je vous prédis que ce qui fonctionne en 2025 ne fonctionnera plus en 2055.

A study indicates that individuals with less knowledge about AI are more receptive to integrating it into their daily lives. It suggests that familiarity with AI may lead to increased skepticism or critical evaluation, whereas limited understanding could result in greater openness to its adoption.

“You can see the computer age everywhere, except in productivity statistics” (Robert Solow). Despite technological advances, productivity gains remain elusive. Is AI breaking this pattern or reinforcing it? Or is our definition of productivity the real issue?

Is digital technology creating a toxic work environment? Lecko's report confirms that hyperconnectivity harms engagement, work quality, and mental health while worsening the carbon footprint. The solution? Better management, as the misuse of technology is not inevitable but a managerial choice.

What if companies looked to freelancers to improve employee experience? After six months as an independent worker, the author identifies three key factors: autonomy (boosting engagement and innovation), flexibility (enhancing productivity), and continuous learning.

Can too much efficiency harm the customer experience? Sukiyabashi Jiro serves 20 sushis in 20 minutes—precision or rush? In service industries, pure optimization can backfire. Automation, rigid workflows, or imposed pace risk disconnecting quality from experience. Where do you draw the line?

Infobesity cripples performance, harms employees, and weakens businesses, yet many see it as inevitable. How do we tackle intangible workflow issues? Proven methods exist, but they require a shift in perspective.

AI is ushering in a new era of human-machine collaboration, creating the Superworker, who is more productive and creative. Adoption follows four stages: assistance, augmentation, automation, and autonomy. Success depends on redefining jobs, skills, and compensation to support this transformation.

SME executives are eager to explore AI but often overlook the regulatory side. It’s like the GDPR era—going from indifference to urgency. To help those unfamiliar with the tech, I’ve summarized the EU AI Act, highlighting key risks and what businesses need to know before diving in.

L’IA générative en entreprise suit trois tendances : ouverture, investissements massifs et spécialisation. Elle s’intègre sous quatre formes : collaborateur augmenté, automatisation des processus, hybridation des tâches et IA autonome.

DeepSeek was inevitable: a low-cost, open-source alternative following China’s usual playbook. It raises a key question—should businesses choose U.S. or Chinese LLMs, or both? Diversifying mitigates risks, but data security and regulation remain key concerns. SLMs may soon disrupt the market further

L’IA générative peine à prouver son ROI en entreprise. ServiceNow propose une approche transverse avec AI Agent Orchestrator, qui coordonne des agents IA cloisonnés dans les silos applicatifs. Son AI Agent Studio, en no-code, permet aussi de créer des agents sur mesure pour fluidifier les processus.

Just read this: interesting, given the "misérabilisme" in some parts of the EU! "Take Nvidia out of the S&P 500, and its total returns underperform the eurozone’s stock benchmark since this bull market began in late 2022."

#ATS filters CVs without intelligence—if a human reads yours, a machine decided so. Many CVs are never seen. ATS favors eligible candidates over high-potential ones, misinterprets profiles, and biases hiring. Before reaching a human, you must first convince a not-so-smart machine.

Businesses must invest in AI despite its high cost and low current ROI. Some are already selling assets to keep up. Sooner or later, the question of profitability will arise. Yet, the decision won’t be based on economic logic but driven by other factors.

GenAI transforms businesses with advanced language processing and content creation, but it’s no strategic oracle. It optimizes and explores scenarios but lacks reasoning and originality. It remind us that humans remain essential to shaping the future :AI is a sparring partner, not a substitute.