AI Audio/Video Gear Now Affordable and Powerful
20 years ago when I worked at Microsoft a researcher showed me a new kind of microphone he built: an array of four microphones built into a box. It cost about $10,000 for him to build. Now @insta360’s new Wave, https://t.co/MokikgunUR an AI speakerphone, has eight microphones, and only costs $255. Plus it has a way better speaker than my Macs have in it. The microphones let me talk to @typelessdotcom (which is an AI that listens and writes into every app and AI I use). And it hears me much better than the other microphones I have. And just got a new robot on top of my Mac that follows me around and has way better video quality than any other camera I have tried. The Insta360 Link 2 is an AI-powered 4K webcam. $150. https://t.co/lArdchPpKj When this camera company launched in my house years ago I had no idea that it would get into selling so many amazing devices to make my life better since I either am on so many podcasts or listening to them, like I am listening to @altryne’s ThursdAI right now, which is the nerdiest show that started here on X audio spaces. All of these are highly recommended.
AI Platform Maps Gene Function at Scale, Redefining Drug Discovery
When I launched insitro , the goal was never to make one drug. If that had been the case, we could have used existing tools. Our goal was to change how drugs are made – and to do that, we...
AI Turns Ideas Into Apps—Join $10K Hackathon
I’m judging the #YouBuildChallenge hackathon (Dec 16–22). The premise is simple: describe what you want to build, and @YouWareAI turns it into a working app. $10K+ in prizes. YouWare is rethinking how software gets created—shifting from writing code to expressing creativity. Built by...
AI‑filled Applications Boost Referrals; Network Now
AI slop in the job application process is only going to increase internal referrals processes and network effects. At least 50% of applications I receive are the same AI answers. To the applicant, they seem impressive and well thought out....
US and China Pursue Different Visions of the AGI Race
Excellent new episode of Your Undivided Attention exposing many more subtle differences between how the US and China think about the race for AGI , and whether they are even competing for the same thing.
Google Open Sources MCP Toolbox for AI‑Driven Databases
Google's MCP Toolbox for Databases is now open source 🔥 A single backend for connections, security, and schema-aware SQL tools, built for AI agents. Compatible with Python, JS, Go, LangChain, and more. Repo in 🧵 ↓ https://t.co/nd3rX7JpFH
Download the Latest Grok App Today
Download the latest @Grok app
Snapshot Engine Enables Reversible Actions for Coding Agents
Coding agents will make mistakes, so they must run on infrastructure that allows every action to be reversed. We built a snapshot engine to achieve this: https://t.co/gdetxGl5z7
Join Codex Team to Accelerate 100x Improvements
Codex is getting extremely good and will rapidly improve. If you want to help make it 100x better next year, the team is hiring. Crazy adventure guaranteed; success probable.
AI Chooses Cells for Costly Immunotherapy via Cellvoyant
ai to select which cells to use for preciously expensive and effective immunotherapy: meet @cellvoyant https://t.co/7SkTFo9QsX
AI Prompting Threatens Microsoft Office's Relevance
Genspark's CEO says "just prompt it." My reaction? Microsoft Office is in trouble as more people discover the new way to work. They won't be going back.
AI-Heavy Jobs Outpace Market in Growth, Wages
AI JOB LOSS HOAX EXPOSED According to a new study from Vanguard, the "occupations most exposed to AI automation are actually outperforming the rest of the labor market in terms of job growth and real wage increases." Job growth was up...
AI Models Now Impacting Cybersecurity with Real-World Vulnerabilities
Last week, a security researcher using our previous model found and disclosed a vulnerability in React that could lead to source code exposure. I believe these models will be a net win for cybersecurity, but we are in the 'real impact...
GPT‑5.2‑Codex Debuts for Agentic Coding Success
GPT-5.2-Codex launches today. It is trained specifically for agentic coding and terminal use, and people at OpenAI have been having great success with it.
2027 AI Race: Alignment Determines Humanity’s Future
🚨 From Agent-1 to Superintelligence: The AI 2027 Scenario The AI 2027 Report outlines one of the most thought-provoking trajectories for artificial intelligence: a rapid evolution from simple assistants (Agent-1) to autonomous, adversarially misaligned systems (Agent-4) and ultimately to Agent-5 —...
Nano Banana
Nano Banana Pro remains my favorite image generation model. Undoubtedly, 2026 will see more AI scams. And more revenue-generating AI influencers. Prompt: This woman is sitting on a white beanbag and laughing. She’s wearing a light pink cap, a black high necked crop...
Humans Savor “Food for Thought”—LLMs Lack Intrinsic Reward
I love the expression “food for thought” as a concrete, mysterious cognitive capability humans experience but LLMs have no equivalent for. Definition: “something worth thinking about or considering, like a mental meal that nourishes your mind with ideas, insights, or issues...
Open‑source PE‑AV Boosts Audio‑visual Separation Performance
We’re open-sourcing Perception Encoder Audiovisual (PE-AV), the technical engine that helps drive SAM Audio’s state-of-the-art audio separation. Built on our Perception Encoder model from earlier this year, PE-AV integrates audio with visual perception, achieving state-of-the-art results across a wide range of...
Google AI Leaders Review Year, Preview Future Innovations
Join us live on X at 1PM PST / 4PM EST to hear @JeffDean, @OriolVinyalsML, @NoamShazeer, and @OfficialLoganK reflect on our year in AI at Google, and explore what's coming next: https://t.co/DnhUg8YYoC
Keras 3.13 Introduces LiteRT Export and GPTQ Quantization
New Keras release: 3.13 🎉 Some major new features: • Model export to LiteRT (formerly TFLite) for mobile/edge • GPTQ quantization support for post-training compression • New Adaptive Pooling layers for dynamic architectures https://t.co/Ogmag7FYCY
Step-by-Step Roadmap to Production-Ready AI Agents
Everyone wants to “build AI agents.” Few have a clear path to learn them. 🧭🤖 This roadmap is a solid, no-fluff sequence for going from zero to production-grade agents: Level 1 — Foundations ✅ Python / TypeScript basics ✅ ML...
End-to-End Neural Network Enables Real-Time Bipedal Gait Control
This is a vertically integrated end to end deep neural network performing forward pass inference real-time, controlling individual actuator's torque output for bidpedal gait generation in adverse, GPS denied, envs. ok its standard PPO rl trained in mjlab, strapped to a...
World Models Poised to Rival LLMs, General Intuition Leads
World models will be as important as LLM's. Next big market and General Intuition has the best data set for this...
Immediate Steps to Guard Against AI Cybercrime
Take These Steps Today to Protect Yourself Against AI Cybercrime AI isn’t only a productivity tool — it’s also enabling new threats. This article guides you through concrete steps to safeguard yourself and your business. Read more 👉 https://lnkd.in/e8Y8Wk3Z #AICybersecurity...
First‑mover Edge Fades in AI and Robotics
idk when first mover advantage was actually a thing, but it sure seems to not be a thing for AI and robotics.
Delfa Builds AI‑first Software for Life‑saving Trials
every industry has its high‑stakes systems: software that cannot fail because the cost of failure is borne by real people. in clinical research, that system is the infrastructure that manages participant journeys through clinical trials. these trials are the gateway...
A16z Argues for Balanced, Innovation-Friendly AI Regulation
The a16z perspective on AI regulation.
AI Outperforms Humans Technically, Humans Still Lead Multimodal Reasoning
AI now exceeds human performance in most technical tasks — but humans still lead in multimodal reasoning 🧠 The next frontier isn’t about replacing humans, but amplifying how we think across data, images & language. Source: Stanford AI Index 2025 #AI #Innovation #FutureOfWork...
Segmentation Is Essential; One Bad Segment Kills Deliverability
Your Emails Are Getting Penalized for This One Mistake (And 90% of Brands Are Guilty) The more you segment by behavioral engagement (opens, clicks, replies), the more inbox AIs will reward you with prime placement, even if your overall list...
Deepnote Open‑Sources Next‑Gen Notebook, Challenging Jupyter
ICYMI, Deepnote, a company I’ve followed and used since its creation, just recently OPEN-SOURCED their modern notebook framework 🔥 And honestly, this could be the final chapter for Jupyter. After 7 years of development, @DeepnoteHQ has built something that redefines the notebook...
Google Launches AI Agents to Automate Dev Workflows
Google is quietly changing how devs work. These 5 AI agents automate GitHub (via Gemini CLI), BigQuery, Looker, Database migrations and more, from data ingestion to PR reviews 🔥 Here’s what each agent actually does 🧵 ↓ https://t.co/0o1Z2jcBia
AI's Future Lies in Embracing Forgetting, Not Accumulation
Why the future of AI depends on what it can let go of and what humans have always known how to forget We talk about intelligence as though it were a process of accumulation. More data. More training. More experience....
Studying Silicon Valley's Failures: Lessons From K-Scale
Silicon Valley is paved with failures. We don't often think about the failures, but it's a hobby of mine to study them and since I was the first to tell you about K-Scale, it's only right that I spend a...
Streamlined Multitasking with Claude via Contextual Chat Streams
Managing a dozen chats with Claude at once ⬇️ Majority of AI superusers I meet with are doing multiple AI tasks at once. Like, 8 different coding windows. Or 16 different client projects. A fun AI hack for me has...
AI-Generated Code Requires More Scrutiny, Has More Bugs
AI-authored code needs more attention, contains worse bugs • The Register https://t.co/5g05bZCTFc #artificialintelligence #aicode #AI #Tech #Coding #100DaysOfCode #codingerror
Innovation Thrives on Strong Links, Not Weakest Links
Innovation is a "strong-link problem". In a chain (weak-link problem), the weakest element breaks the system. In discovery (strong-link problem), the strongest element makes the breakthrough. The rest of the system provides the infrastructure that allows the outlier to function
World Models: The Path to True AGI
Why World Models are the way to AGI.
AI Will Tackle Complex Group Dynamics in Business, Politics, Economy
After math, code, assistant competence AI will take on complex group dynamic problems like firms, politics, economy.
ChatGPT Launches App Directory for Developer Submissions
🚨 JUST IN → Devs can now submit their apps to ChatGPT. "Approved apps will be listed in the app directory, which is a new surface for users to search for apps directly in ChatGPT" https://t.co/KLyf91GOhk
Our Brains Are Tuned to the Universe’s Stable Laws
Because our universe follows stable laws, a sufficiently general intelligent system adapted to it, like human-driven science, can eventually model any phenomenon within it. Human intelligence may not be "universal" in the mathematical sense (see No Free Lunch theorem), but we...
First to Build Grok Voice Agent for Reachy Mini?
Who's going to be the first to build a grok voice agent app for reachy mini?
Gemini 3 Flash Masters Multimodal Geometry Calculations
In this demo, you’ll see Gemini 3 Flash’s frontier-level reasoning and multimodal capabilities on display. The model is able to simultaneously conduct complex geometric calculations while processing complex inputs (video and image). You can play around with the slingshot in...
AI Marketing Is Just Evolution, Not Revolution
If you've been doing SEO, running social ads, or optimizing for any algorithm, you've already been doing AI marketing. The panic around AI agents and machine learning is overblown because marketers have been marketing to machines for over a decade....
AI Triggers Cascade: From Efficiency to Industry Overhaul
AI-driven shifts are reshaping productivity, market behaviors, competitive pressure, and organizational design, creating a sequence of changes that move from immediate efficiency gains to broader transformations across industries. Infographic @Gartner_inc via @antgrasso https://t.co/BckFf1DZ02
AI‑Powered Cloud Browser Redefines Multi‑Site Web Interaction
How is the Web changing due to AI? This visit to @Tiny_Fish yesterday goes deeply into just how. TinyFish makes a new kind of web browser. One that is started virtually. It is like a Google Chrome in the cloud. But it...
Agentic AI & MCP Power Sustainable Enterprise Transformation
🚀 The Future of Enterprise Transformation Starts Now 🔍 Harnessing Agentic AI and Model Context Protocol (MCP) for Sustainable Growth 🔹 Businesses are entering a new era driven by autonomy, security, and intelligent adaptability. 🔹 Agentic AI and MCP are redefining...
Run Fine‑Tuned LLMs Directly on Your Phone
Wow. You can now fine-tune LLMs and deploy them directly on your phone 🤯 @UnslothAI × PyTorch makes it possible to export and run your Unsloth AI trained model → 100% locally → on iOS + Android eg deploy Qwen3 on Pixel 8 & iPhone...
AI Data Centers Don't Actually Use Massive Water
The claim that AI data centers consume a large amount of water is a hoax.
Nvidia's Nemotron 3 Tops HuggingFace, Boosts US AI Leadership
Nemotron 3 is already third trending on @huggingface. @nvidia is really becoming the American powerhouse for open-source AI! https://t.co/uDQWBPJUwL
Gemini's New Fast Model Reinforces Known List Benefits
Good morning. Gemini has a good new model out that's fast. But you already knew that if you had this list up and running: https://t.co/wAjs9SAZfe
