Bruno Herrmann

Physical Intelligence, a hot robotics startup, says its new robot brain can figure out tasks it was never taught techcrunch.com/2026/04/16/phy… via @techcrunch

Physical AI is not coming. It has already come in China. From factory floors to front desks: Top humanoid robot makers to watch jingdaily.com/posts/from-fac…

Uber CTO Shows How Claude Code Can Blow Up AI Budgets — The Information theinformation.com/newsletters/ap…

How France’s Mistral Built A $14 Billion AI Empire By Not Being American via @forbes forbes.com/sites/iainmart…

Beyond the Black Box: the new ‘explainability’ rule for enterprise AI raconteur.net/technology/bey… via @raconteur

AI isn’t replacing us. It’s convincing us we can replace each other buff.ly/RqfJWV4

"Should there be some centralized government agency that's set up to make sure the latest model that comes out is safe and secure?" says @RepJoshG on regulating AI. "Some of my colleagues think we should try to put it back in a box. That's ridiculous." cnb.cx/4cSW2DV

Map of Norwegian Dialects Larger zoomable image here: pbs.twimg.com/media/HG39JERb…

Singapore’s hunt for AI stars risks alienating small players techinasia.com/singapores-hun…

On @CNBC this morning: To be or not to be an AI champion? Our Director General @BonefeldCecilia warned that Europe risks undermining its AI ambitions without targeted simplification in the #AIOmnibus. Overlapping rules could mean double compliance costs: up to €600k per SME per product and €31bn in year 1 across the EU economy. Europe needs a coherent, streamlined framework that supports AI innovation and competitiveness. Watch the full interview 👉 bit.ly/4cMTQO1 #SimplifyForStrongerEurope

How agencies are reshaping junior-level roles and training in the AI era adage.com/agencies/aa-ne…

OpenAI has long been the leader for paid usage by U.S. businesses, but Anthropic has closed the gap with tools like Claude Code and Cowork 🤖 This graphic is part of AI Week, our @TerzoHQ-sponsored series. It explores the share of U.S. businesses paying for AI models from different providers, with the data from Ramp. visualcapitalist.com/ranked-ai-mode… #aiweek

EU AI Act Compliance: a technical audit guide for the 2026 deadline raconteur.net/global-busines… via @raconteur

AI use across Europe varies widely, from 56% in Norway to about one-third in the UK and Germany 💻 This graphic is part of AI Week, our @TerzoHQ-sponsored series exploring how artificial intelligence is reshaping business, technology, and society. visualcapitalist.com/mapped-ai-adop… #aiweek

Anthropic Labs Launches Claude Design Tool for Visual Prototyping vktr.com/digital-market…

Elon Musk's AI company makes ludicrous job posting sfgate.com/tech/article/m… via @SFGate

Flood of small parcels from China pushes Belgian airport to limit sc.mp/s0irg?utm_sour… via @scmpnews

AI Is Recreating a Colonial World Order znetwork.org/znetarticle/ai…

⬛️ The Smartest AI Models of 2026 The ranking offers a snapshot of how today’s leading AI models perform on abstract pattern-recognition tasks, and just how close the race has become. ▪️Why This Benchmark Matters TrackingAI’s leaderboard is useful because it offers a simple, familiar way to compare reasoning performance over time. The site [tinyurl.com/mrywck6z] also notes that if a model refuses to answer, it is asked the same question up to 10 times, and the most recent answer is used for scoring. ▪️What This Benchmark Does Not Measure Still, an IQ-style benchmark captures only one slice of capability. It does not measure everything that matters in real-world AI use, such as coding ability, factual reliability, tool use, or performance in professional domains. An IQ‑like benchmark can be useful as a limited tool for measuring reasoning ability. However, referring to it as an “IQ” is conceptually incorrect, professionally reductive, and communicatively misleading. It does not measure an overall “level of intelligence,” but rather a specific, context‑dependent performance dimension - and it should be labeled and interpreted precisely as such. 📢 Why Calling This an “AI IQ” Is Misleading What this chart really shows is not “intelligence”, but a narrow slice of abstract reasoning performance under artificial constraints. Useful as a datapoint - dangerous as a narrative. ✨ @Khulood_Almani @AkwyZ @MaryRich78 @rwang0 @timo_vi @drsharwood @TamaraMcCleary @DrHolzwarth @HelenBevan @phinifa @pierrecappelli @JimHarris @jenstirrup @GlenGilmore @subare @Ronald_vanLoon @enilev @Scobleizer @AndrewYNg @YuHelenYu ✨ #ArtificialIntelligence #AIReasoning #AIBenchmarks #DigitalTransformation #People #ModelEvaluation #AI_IQ #AIResearch #MachineLearning #ML