My latest piece on @FT about the Chinese LLM open-source advantage. Contrasted with the secretive development of LLMs in the US competing in a winner-takes-all battle to reach AGI, Chinese models offer a distinct pathway for progress in AI when they are forced to play catch-up. Through open technologies, the strategy might lead to a broader global reach.
I agree with this perspective: the performance of an AI system depends far more on how the agent is constructed, tuned, and evaluated than on the specific foundation model behind it. Tools like the Opik Agent Optimizer go beyond manual...

The Generative AI ecosystem is evolving into a full tech stack — powering autonomous AI agents. From infrastructure and LLMs to RAG pipelines, agent behaviors and orchestration layers, this framework shows the 6 layers driving next-gen AI systems. Credit: @goyalshalini #AI #GenerativeAI #AgenticAI...
AI now can start a virtual web browser, and your AI agents can use it like a human would, getting data out of the Web that was hard to get to before.
Take a photo with @Grok. It can analyze anything.
This keeps happening with Claude, and the reason it matters is simple. Once several tools in a category clear the quality bar, reliability decides which one people keep open. You feel it when something real is underway and time matters. These tools...
My teammate built me a Bosnian language learning app using Gemini and now I have no excuse not to be fluent by next month. https://t.co/Dc0iERR4eM
The front end of commerce resists automation because understanding forms before decisions. Generative AI changes that dynamic. 1. Buyers express intent in natural language 2. Discovery unfolds through open questioning 3. Alignment forms before commitment This is how intelligence reshapes...
How Zoom Is Revolutionizing Work with Agentic AI — From Meetings to Milestones AI powered tools from Zoom are moving beyond meetings, helping teams track progress, automate tasks and reshape how we collaborate at work. Read more 👉 https://lnkd.in/e_8_DbgW #AI...
AI-focused chips offer tailored processing for deep neural networks, improving speed, energy use, and sector-specific capabilities while supporting new advancements that guide the evolution of intelligent systems across devices and industrial domains. Microblog by @antgrasso https://t.co/QKYoX0OdN9
Humachineology: Design Orientation and Systemic Implications A distinguishing feature of humachineology is its design-oriented approach to shaping the interaction and integration of humans and machines. This perspective views the future not as a deterministic outcome of technological progress but as subject...
Sharing content online is one of the highest-ROI habits you can build, and it has nothing to do with going viral. When I started, my goal wasn’t audience or money. It was learning. Content forces clarity. Saying “I want to learn...
Stop Blasting Everywhere. Start Building a Content Ecosystem. The more you slice one pillar piece into native, platform-specific content, the more you dominate YouTube, TikTok, LinkedIn, and even AI search, without burning out. The sooner you realize AI is increasingly...
The 2025 Artificial Intelligence and Business Strategy report, from @mitsmr and @BCG , looks at how organizations that are adopting agentic AI are gaining advantage while facing four distinct tensions. Read the full report >> https://t.co/yt7MtRuGRb #CES2026 #AI #IoT...
Humachinekind refers to the species of humans and machines living, working, and evolving together.
A proud moment to share. Congratulations to Team DementAI on an outstanding achievement at the #SASHackathon With more than 55 million people worldwide living with dementia, earlier detection and better awareness are critical, not only clinically, but socially and economically....
AI churn has IT rebuilding tech stacks every 90 days | CIO https://t.co/DGHKBQznf6 #techstack #TechRevolution #AI #ArtificialIntelligence
Grok Imagine is radically better now
Oh My! I just built a new RL-native system with 2.6k parameters that matches a specialized 62B model from a 2024 paper, specifically on GSM8k. That's over 20,000,000x smaller than the baseline — but the approach is so different the traditional...
8 Smartphone Trends That Will Shape 2026 #Smartphones in 2026 will bring major upgrades in #AI , connectivity and design, creating devices that feel smarter and more capable across work and personal use. This article explores the #trends that will...
Try @Grok voice mode. It’s the best.
Build vs buy is dead — AI just killed it | @VentureBeat https://t.co/PxNyqvZAYc #AI #ArtificialIntelligence
Are your AI agents still stuck in slideware or earning trust in production? At #AWSreInvent, I spoke with Scott Clark, CEO of @dbnlAI, an @nvidia Inception startup, about “agent analytics”, the 3rd pillar of enterprise AI: build, run, understand. Watch our full...
Can robots help dementia patients? https://t.co/6MoqJ88qly #dementia #HealthcareAI #healthcare #AI #Robot #humanoid @SpirosMargaris @PawlowskiMario @mvollmer1 @gvalan @ipfconline1 @LaurentAlaus @Shi4Tech @Fisher85M @kalydeoo @Ym78200 @Nicochan33 @chboursin @3itcom @Fabriziobustama @sallyeaves @helene_wpli @ahier @rwang0 @EvanKirstel @RLDI_Lamy @Analytics_699 @Khulood_Almani @tewoz @chidambara09 @IsabellePiel29 @ClementIsa @SvetBnov @mallys @jeancayeux @aure79lien...
Speed-of-Light Computing Could Get Us to AGI Very Fast https://t.co/LeBQSYsO4w #photonic #AI #ArtificialIntelligence #chip @AlbertoEMachado @Eli_Krumova @postoff25 @Khulood_Almani @anand_narang @NutritiousMind @baski_LA @TanyaSinha_ @devaang @AlAmadi1 @jeancayeux @enilev @efipm @mvollmer1 @Nicochan33 @RagusoSergio @FrRonconi @Shi4Tech @sallyeaves @LaurentAlaus @Fabriziobustama @smaksked @MargaretSiegien @PawlowskiMario @gvalan @Ym78200 @mikeflache @EduardoValenteI...
I spent last week at Neurips (the big ML/AI conference) and had hundreds of amazing technical conversations with old friends and new. Beyond the technical, I'm deeply grateful for the love from all the folks who stopped me in the...
Prompt engineering is now a critical skill for getting the best results from ChatGPT. These 17 techniques — roles, context, structure, scenarios, frameworks and more — can transform the quality of outputs. Credit: @ginacostag_ #AI #ChatGPT #PromptEngineering #Productivity https://t.co/6khSHbgNG1
in jan ‘25, uk pm starmer pledged to mainline ai into the veins of the uk with the ai opportunities action plan not only has gov been very slow to deliver on much of it it seems the pm has missed...
But you can't do self driving without LiDAR. So many "experts" told me that over the years. I was given so much shit for telling them they were wrong. I guarantee none will apologize.
Grok says: "In summary, watching this interview will equip you with a comprehensive, insider's view of why 2026 could be a breakthrough year for decentralized AGI, backed by technical, philosophical, and practical arguments. It's particularly valuable for AI enthusiasts, developers, policymakers,...
The road to AGI with @bengoertzel. Very special conversation with one of our pioneers. Evaluates new technology coming in 2026. https://t.co/R8lN0DYR2Z
How I explain AI to business teams in 30 seconds 👇 This infographic is a great map of the landscape, because it shows AI in three clear layers: 🟩 Traditional AI Predicts, classifies, and detects anomalies. Think: demand forecasting, fraud...
If you are interested in understanding the design and components of modern LLM architectures, I have extensively grown and updated the Big Architecture Comparison article I published last summer. It grew 2x in size since then: https://lnkd.in/g-dwdPqy 1. DeepSeek V3/R1...
Massive infrastructure move: @OpenAI signs ~$300B deal with @Oracle for cloud compute over ~5 years (Project Stargate), starting 2027. That kind of scale reshapes how AI systems are built. 🔗 https://t.co/tiNoxHT02y #OpenAI #Stargate #CES2026 #AI #IoT #5G https://t.co/Fk15YtbFXj
AI governance is not optional, it's essential. The EU Artificial Intelligence Act categorizes AI risk to protect rights and safety. Discover more with @ingliguori. #AIAct #Governance #ModelOps #AIethics https://t.co/Htivyq426P
I’ve been watching a pattern lately: a lot of businesses are trying to “add AI” hoping it will magically fix everything. But here’s the honest truth: AI won’t save a bad business. It will simply reveal what’s already broken. If your operations are messy,...
two new essays on @airstreetpress next week: 1) tues: can ai generate new science? on ai agents for discovery 3) thurs: rebuilding high-stakes software ai-first - a case study in clinical trials sign up at press(dot)airstreet(dot)com
AI investment is increasingly concentrating in a limited number of nations, with the United States and China standing out as primary points of gravity for capital, talent, and strategic direction. From these national ecosystems emerge the conditions that support scale,...
This is a big. Train LLMs ~3× faster with no accuracy tradeoff. @UnslothAI just rolled out a strong set of kernel + data pipeline upgrades that remove real training bottlenecks: padding overhead, unfused ops, and long-context indexing pain. Result: models like Qwen3-4B now...
I just ran one of these prompts. Asked Grok about you. Here is its answer about what you desire, fear, and obsess over: - **Desires**: Early access to cutting-edge demos, prototypes, and announcements in AI, robotics, AR/VR/spatial computing, autonomous vehicles, and brain-computer...
I appeared on Bloomberg to discuss President Trump’s new Executive Order supporting a national standard for AI regulation. Thank you @business @EdLudlow @CarolineHydeTV https://t.co/9Q6foJE5ZK
Google AI Edge: Google's Advantage Over ChatGPT Unfair? https://t.co/tqNQnpD5RW #Gemini #openai #LLMs #GenAI #AI #generativeAI #ArtificialInteligence
Enabling small language models to solve complex reasoning tasks | MIT News | Massachusetts Institute of Technology https://t.co/4ykZwiTvuN #LLMs #LLM #SLM #GenerativeAI #GenAI #ArtificialInteligence
In 2018, Ron Schmelzer and I set out to solve a practical problem: how do you manage AI projects responsibly and deliver business value? That became CPMAI, a vendor‑neutral methodology designed for real‑world AI delivery. ➡️ Fast forward to today:...
Confused about the different competing views around whether A.I. will be good or bad that don't seem to acknowledge each other? "Faces of AI," is a short video that aims to synthesize seemingly opposing perspectives on AI: https://t.co/tEXlavoJ9v It's part of a series...
Cheat sheet: ChatGPT = create/build (writing, coding, workflows) Grok = live trends + punchy takes Gemini = Google Workspace-native collaboration Claude = deep reading + long-doc reasoning Perplexity = research w/ citations Right tool for the right job. #AI #LLM #Productivity https://t.co/yIkFj6w8T4
Every December, I deliver “Next Year in AI Leadership” talks to my clients. And this is the first year where every single slide feels like sci-fi. Like, what do you mean I built a sign language app in <20min? What...
I've been researching connections between water, spirituality and GenAI for around two years. The Introduction and Abstract below represent my first crack at developing key themes I'm working on with friends and thought leaders including Ben Christensen , Nicoleta Acatrinei...
AI agents aren’t powered by “one LLM.” They’re built by choosing the right model type for the right job. This infographic nails it: 8 LLM families show up inside modern agentic systems, each with a distinct role: 🔹 GPTs –...
Is this AI erotica announcement happening? It’s December. Sam said December. Or will it be smuggled in again? Or was the backlash enough to delay or stop the rollout? Upon your first adult prompt, do you have to confirm your...