Today's AI Pulse
OpenAI launches ChatGPT Images 2.0 with multilingual text and real‑time web research
OpenAI has introduced ChatGPT Images 2.0, a next‑generation image model that adds multilingual typography, real‑time web research, and agentic reasoning to its generation pipeline. The new “Thinking” mode lets paid users produce up to eight coherent images per prompt, including floor plans, maps, UI mock‑ups, and apply capabilities to uploaded content.
Also developing:
China Wants to Sink Data Centers Underwater - Could This Be the Next Frontier in Computing?
China is launching ocean-cooled data centers near Shanghai, using renewable energy to cut cooling costs and advance sustainable computing.
AI Is Unlocking Rapid Personal Reinvention for Anyone
Between Sora 2 and Claude Imagine, it’s clear we’re heading toward a world where anyone can learn anything (and I’d even argue we’re already there). The potential is enormous and so few are seizing it. Most people use AI to rewrite...
AI Lives in Massive Data Centers, Not Just the Cloud
We got to tour a cutting-edge data center. Here's what we learned. "Every AI response you see is powered by a hidden backbone — billions of dollars of massive warehouses cooled by air, sometimes water, with enough electricity to run...
Google’s Photoshop-Killer AI Model Is Coming to Search, Photos, and NotebookLM
Google is integrating its Nano Banana image‑editing model from Gemini 2.5 Flash into Search (Lens and AI Mode), Google Photos, and NotebookLM, letting users perform conversational image edits and apply new Nano Banana–powered video styles directly in those apps. The...
Master AI in 2025: From Basics to Career
🚀 Master AI in 2025 Dal codice alla carriera 👇 📘 Build strong foundations 📗 Master ML & Deep Learning 📙 Learn LLMs & MLOps ⚙️ Deploy real AI projects 💼 Transition to AI careers L’AI non è più un’opzione — è una competenza chiave. Qual è il...
Altman: Early Sora Release Prepares Society for Deepfakes
I asked Sam Altman if he's worried about watermark removers on Sora. Everyone's seen the unfiltered Jake Paul videos with no watermark (generated by Sora) going viral on socials... So what's the pull for creators to enable face cameos with that risk? Sam's...
OpenAI Partners with Broadcom to Produce Its Own AI Chips
OpenAI is teaming up with Broadcom to produce its own computer chips to power its AI data centers. The deal is the latest in a series of partnerships designed to reduce the company’s reliance on Nvidia and secure enough computing...
Run Your Own ChatGPT Clone in Just Four Hours
Excited to release new repo: nanochat! (it's among the most unhinged I've written). Unlike my earlier similar repo nanoGPT which only covered pretraining, nanochat is a minimal, from scratch, full-stack training/inference pipeline of a simple ChatGPT clone in a single, dependency-minimal codebase....
California Becomes First State to Regulate AI Companion Chatbots
California has enacted SB 243, becoming the first state to regulate AI companion chatbots, aiming to safeguard children and vulnerable users from potential harms. The legislation introduces stringent guidelines for the development and deployment of these technologies, mandating transparency and...
A New Wave of 'Shadow AI' Is Worrying Workplaces
New Microsoft research shows that the use of shadow AI is once again on the rise.
AI Defies SaaS Rituals; Embrace Hypothesis over Fixed Roadmaps
SaaS let you win by drawing straight lines. Requirements in, code out. QA ran their checklist. Customers got exactly what the roadmap promised. The whole machine ran on one idea. If you could predict it, you could control it. AI ignores your...
This New AI Technique Creates ‘Digital Twin’ Consumers, and It Could Kill the Traditional Survey Industry
A new research paper quietly published last week outlines a breakthrough method that allows large language models (LLMs) to simulate human consumer behavior with startling accuracy, a development that could reshape the multi-billion-dollar market research industry. The technique promises to...
Leaders Own AI Mistakes: Scale Your Decisions Wisely
❗ AI doesn’t make mistakes — it scales yours. When AI fails, it’s not the algorithm “misbehaving.” It’s leadership. How you build it is riddled with choices leaders must own. Even when you buy off-the-shelf AI, it's the leader who...
YouTube Music Is Channeling Spotify’s AI DJ with This New Feature – and Early Testers Love It Despite One Big...
Despite some frustrations, beta users are enjoying YouTube Music's new AI host – especially its music trivia commentary.
Import AI 431: Technological Optimism and Appropriate Fear
What do we do if AI progress keeps happening?
Three Learning Pillars Power Modern AI Systems
🔹Supervised. Unsupervised. Reinforcement. Three pillars of #MachineLearning — three ways AI learns: 📘 Supervised → learns from labeled data 📗 Unsupervised → finds hidden patterns 📙 Reinforcement → learns by trial, error & reward Each builds the foundation of intelligent systems. Which one do you think...
Salesforce Announces Agentforce 360 as Enterprise AI Competition Heats Up
Salesforce announces an upgraded version of its Agentforce platform designed to help enterprises build and deploy AI agents.
Slack Is Turning Slackbot Into an AI Assistant
Slack is testing an update for Slackbot that transforms it into an AI assistant. Presently, it operates as a tool for delivering reminders and notifications. With the update, Slackbot can create custom plans tailored to your workspace, sift through a...
Anduril’s New EagleEye MR Helmet Sees Palmer Luckey Return to His VR Roots
Anduril Industries has unveiled its new EagleEye MR helmet, marking a significant return to virtual reality for co-founder Palmer Luckey. This advanced mixed reality headset is designed for military applications, integrating augmented and virtual reality capabilities to enhance situational awareness...
Rude Prompts Boost GPT-4 Accuracy, Politeness Hurts
Being rude to AI actually makes it smarter! A new study, “Mind Your Tone”, tested how prompt politeness affects GPT-4-based models. The result: very rude prompts scored 84.8%, while very polite ones dropped to 80.8%. Statistical tests confirmed it’s not random. Politeness consistently...
AI's Self‑Fueling Loop: Compute, Demand, Investment
The AI money loop! @Nvidia feeds compute, and invests $100B straight into OpenAI. @OpenAI, valued at $500B, powers growth through $300B in Oracle cloud and AMD GPU scale-ups. @Microsoft, @CoreWeave, @Intel, and @NebiusAI expand reach. @xAI, @Mistral, and Figure AI close the loop through...
From Tool Sprawl to Defensible Value in AI for Legal - with Christo Siebrits of AbbVie
In this podcast episode, Christo Siebrits, AbbVie’s senior associate and general counsel, explains how legal teams can transition from a chaotic array of AI tools to a defensible, value‑driven AI strategy. He emphasizes safe generative‑AI adoption through human‑in‑the‑loop workflows, careful...
2026’s Top Five Tech Trends Shaping Business
Thank you for reading my latest article The Top 5 Technology Trends For 2026. Here at LinkedIn and at Forbes I regularly write about management and technology trends. To read my future articles simply join my network by clicking 'Follow'....
Breaking the Bottleneck: Why AI Demands an SSD-First Future
Presented by Solidigm As AI adoption surges, data centers face a critical bottleneck in storage — and traditional HDDs are at the center of it. Data that once sat idle as cold archives is now being pulled into frequent use to...
Your Standards, Not AI, Determine Output Quality
AI doesn’t care about the quality of your work unless you do. Most people drop in a prompt and expect miracles. But the output never rises above the discipline of the person using it. If you accept average, so will your tools....
Speak, Let AI Polish: Faster, Cleaner Writing
I was so honored when @RandyWKirk1 asked me to write the foreword of a book he, and a bunch of other AI luminaries is writing. I turned it in this morning and he wrote back "Huge thanks Robert. That was...
👾 LLMs Create Token... Humans Create Value
Why human insight, judgment, and creativity remain the ultimate edge in an AI-driven, token-generating world.
Bought a $499 Virtual Dog, Feeling Stupid
On vacation I thought of one startup more than any other: @IllusionOfLife. I hear I'll be one of the first to get its product: a virtual dog. No other company has caused me so much anxiety before getting its product in...
AI's Gravity: Balancing Optimism with Necessary Fear
Technological Optimism and Appropriate Fear - an essay where I grapple with how I feel about the continued steady march towards powerful AI systems. The world will bend around AI akin to how a black hole pulls and bends everything...
Nvidia’s AI Empire: A Look at Its Top Startup Investments
Over the last two years, Nvidia has used its ballooning fortunes to invest in over 100 AI startups. Here are the giant semiconductor's largest investments.
Let AI Automate Tasks, Save Hours of Writing
You should be a bit more lazy. Stop doing things you can automate. Let AI help with that... Here's how you can create and give tasks to your team that go in-depth and cover all the bases without requiring 3...
Neural‑Band‑Free Interaction Boosts Gaming, Learning, AI
neural band, without the neural band this mode of interaction + genie3/veo3/odyssey will be incredible for gameplay, learning, embodied ai training and entertainment https://t.co/nNtV8xLKS7
Strange AI-Generated Clips of Musicians with Animals Are Going Viral
It's fake. And odd. If nothing else, generative AI has proven useful for people who want to make odd stuff for no real reason. The latest thing? Musicians accompanied by animals. Accompanied in the official, musical sense, and also in an even...
Algorithms Now Compose Music for Spotify's Playlists
Spotify made music for algorithms. Now algorithms make music for Spotify. https://t.co/KogbvVIwo4
Sliding Window Attention Streamlines Long-Sequence Processing
Sliding Window Attention 🔗 https://t.co/LIR9gCActT https://t.co/NSn4D4vLoF
When Dirt Meets Data: ScottsMiracle-Gro Saved $150M Using AI
How a semiconductor veteran turned over a century of horticultural wisdom into AI-led competitive advantage For decades, a ritual played out across ScottsMiracle-Gro’s media facilities. Every few weeks, workers walked acres of towering compost and wood chip piles with nothing more...
Real-Time Interactive AI Arriving Within Two Months
It’s time for real-time. Wonder if it’ll be real-time video or *interactive* real-time generation i.e. genie 3
Add .md to URLs for Instant Markdown Access
cool new feature from huggingface. appending ".md" to any docs url will now give you the markdown which can be used as context for your favorite llm or more 🚀 https://t.co/BlRVWjRp4R
OpenAI Releases Free Prompt Packs for Every Role
OpenAI just dropped `Prompt Packs` ↳ 300+ curated prompts ↳ 12+ departments (IT, HR, Sales, Product, CS…) ↳ 25 per role Adapt them with your KPIs & product context for better results! 🔥 100% free. Link to the packs in 🧵 ↓ https://t.co/SBWAfWfpG3
Self‑Driving Tech Already Outdrives Human Drivers
Autonomous vehicles are passing humans in capabilities. And they are way way way smoother at driving than you are. And if you think that is not a true statement it will be by the middle of next year for sure.
Balancing AI Optimism with Cautious Fear in Personal Essay
Gosh I hope none of my overly raw, very personal, and emotional essay about my relationship to AI, called "Technological Optimism and Appropriate Fear" gets misinterpreted! (Comes out tomorrow AM ET in Import AI). https://t.co/aovFZqjxpg
AI Still in BlackBerry Era, Human Review Crucial
FROM THE LIVE SHOW: Wade Foster (@wadefoster), CEO of @zapier, on AI's "iPhone moment." "We're kind of in the BlackBerry era of AI agents. There's a lot of work to get us to that iPhone moment." "Middle to middle is a...
From Months to Weeks: IBM Cuts AI Rollout Time
8 years ago: Implementing a chatbot took months. Our PhD data scientist would wait hours for each data cycle. Moving from concept to production felt like an eternity. Today: NDS Cognitive Labs delivers AI proof of concepts in 2 weeks. That's the...
Agentic Context Engineering Boosts AI Performance Without Fine‑Tuning
🚨 BREAKING: @Stanford just proved you don’t need to fine-tune an AI model to make it smarter 🧠 ACE = Agentic Context Engineering ↳ Model writes, reflects & edits its own prompts ↳ Learns from wins and failures ↳ +10.6% over GPT-4 agents w/...
AI+, Connected Industries Drive Smarter, Secure Enterprise Networks
👉 These themes - AI+, Connected Industries, and Connected Enablers - speak directly to the questions every enterprise leader is asking. #MWC25 Las Vegas #AI #5G #IoT #AgenticAI @MWCHub @CTIA @GSMA https://t.co/st3b04wHGj
Stanford's BLAST Gives LLMs Lightning‑Fast Web Browsing
Bring web browsing to your AI stack, without the latency. @Stanford just dropped BLAST, a high-performance engine that lets LLMs browse, extract & stream the web fast ⚡ ↳ OpenAI-compatible interface ↳ Automatic parallelism + caching ↳ Real-time streaming Free repo in 🧵↓ https://t.co/xTUtwgITv2
MIT Tool Lets You Visualize and Edit Impossible Objects
MIT tool visualizes and edits “physically impossible” objects https://t.co/0lHT0TIJ9m @MIT #AI #IoT #5G https://t.co/uYLG8MyYK6
Reachy Mini Wins TIME’s Best Invention, Targets Million Units
So proud to see Reachy Mini named one of the Best Inventions of 2025 by @TIME! Huge credit to the @pollenrobotics and @huggingface teams, turning a concept into thousands of units sold and shipped in under 6 months. We might not be...
Free AI Courses From Leading Tech Companies and Universities
Here are free AI classes you can take online from top tech firms, universities https://t.co/ibbAfx1Jqf #AI #IoT #5G @FortuneMagazine https://t.co/IKQmuIU2SZ
India’s AI Surge Vital for Global Future
Today I met with PM @narendramodi to discuss Anthropic's expansion to India—where Claude Code use is up 5× since June. How India deploys AI across critical sectors like education, healthcare, and agriculture for over a billion people will be essential...