AI Must Master Tiny Scientific Loops Before Curing Cancer
Many people expect that current AI is ready to cure cancer and do breakthrough new science. ARC-AGI-3 envs are like a microcosm of the scientific method: you must observe a tiny world, form a theory of how it works, test it, iterate until correct. Over the course of a few minutes. If AI can't do it in an ultra-simple, ultra-small scale setting that is explicitly designed to be as accessible as possible, I expect there are a few steps missing until AI can crack the nature of reality.
Reflection AI Seeks $2.5B to Build Sovereign AI Hub
I got to reflect on @reflection_ai 😉 Reflection AI aims to counter Chinese AI systems by promoting freely available, high-performance U.S. AI models. As of March 25, 2026, Reflection AI is in talks to raise $2.5 billion at a $25 billion...
Granola Co‑founder Discusses AI’s Meeting Future
Granola is one of the few AI apps I can't live without. We had co-founder Sam Stephenson join the podcast this week. We discuss the future of Granola and AIs in meetings, the SF vs London tech scene, why...
Physical AI and Robotics Spark Next Industrial Revolution
Physical #AI and #Robotics will bring about the next industrial revolution by @nvidiaomniverse #Innovation #EmergingTech #Tech #Technology https://t.co/k57mXQkVkE
Energy‑Per‑Prompt Redefines AI Query Cost Metric
#Technology #Thread #Semiconductor #Manufacturing #Energy The Semiconductor Energy Per Prompt: 1/ - Energy Per Prompt Defines How Much Electrical Energy Is Consumed To Deliver One Meaningful AI Response. - It Shifts Focus From Peak Power Or Total Datacenter Energy To The Cost...
Teaching AI Irony Turns Its Replies Into Laugh Tracks
AI Ethics: We tried to teach an AI 'irony' by making it watch 10,000 hours of 80s sitcoms. Now it just answers every question with a canned laugh track. 📺😂 #AIEthics #SitcomLife #DigitalIrony https://t.co/O4KtDbM0mx
AI Automates Radiology Tasks, but Radiologists Still Thrive
On Lex Fridman's podcast this week, Jensen Huang brought up radiology and AI, highlighting radiology as an important data point for how we can expect AI to influence jobs. He talked about how nearly a decade ago, AI researchers predicted...
AI-Generated Emails Strip Away My Authentic Voice
After a week of AI-written emails, I noticed something unsettling — I didn’t quite sound like myself anymore. https://t.co/nJ1kEbO1VH
AI Hype Masks Executive Incompetence, Warns Nvidia Crash
Executives use the promise of an AI revolution to wallpaper over their managerial incompetence. https://www.benzinga.com/markets/equities/26/03/51474651/nvidia-could-crash-over-70-warns-scott-galloway-ai-corporate-ozempic
AI Tools Double Top Teams' Output, Automate Routine Coding
AI tools are now standard in software engineering, doubling output for top teams while autonomous agents take on routine coding. https://t.co/AaTgXwZDeR
Ordinary Testers Solve ARC‑AGI‑3 Tasks without Training
To be clear, all ARC-AGI-3 environments are feasible by humans with no prior ARC-AGI-3-specific training. Our bar for feasibility is the following... Each environment was seen by 10 human testers. If 2 testers could independently clear it (successfully solving *all* levels...
Centralized AI in Slack Turns Knowledge Silos Into Compound Power
Your team's ChatGPT tabs are silos. One person knows something, the rest start from zero. I run AI in Slack. 13 channels. One agent sees product decisions, technical builds, strategy threads, everything. Knowledge compounds instead of resetting. That changes what's possible....
AI Thrives on 5G, Edge, Secure Network Partnership
AI adoption isn’t just about models; it’s about infrastructure. Integrated 5G, edge computing, and secure network architecture are enabling distributed intelligence across manufacturing, logistics, and energy environments. New partnership announcement: https://t.co/b665Yqudzw @TMobileBusiness Partner
AI Value Emerges when Embedded And
RT AI value doesn't live in slideware or labs. It shows up when agents are embedded in day-to-day work, governed like core systems, and held to the same standards as any critical operation. #AgenticAI #EnterpriseAI @Star_CIO https://t.co/VtuS5rQ0ZG
AI Coaching Democratizes Continuous, Data‑Driven Talent Development
JUST IN: Democratizing access to professional coaching could alter the talent development paradigm, shifting the focus from occasional, high‑cost interventions to continuous, data‑driven support. https://www.pulse.bot/human-potential/news/huckleberry-rolls-out-20permonth-ai-coach-to-democratize-employee-development-bea5b053-5434-4aeb-a827-679e88ecd41c/ SaaS
AI Rewrote Canon Webcam App in Rust, Fixing Crashes
Great little story from @danshapiro about how he asked a coding agent to fix the official webcam software from Canon that kept crashing. He woke up to a new, fully functional Rust webcam app that has worked ever since. ...
AI Smart Glasses Secure Funding for Dementia Care, Efficacy Unclear
AI smart glasses win major funding to support dementia care, though real-world effectiveness and long-term impact remain uncertain despite early promising observations. https://t.co/vX7jBV2aJt
AI Automates Podcasts, but Loses the Human Spark
"At first I thought you were sick. Not yourself. Low mood. Then I was convinced you were reading a script. By the end I'd already twigged before you even said it was AI." "Your personality is the best thing about your...
ServiceNow Aims to Own Enterprise AI Governance
Everyone is obsessing over who has the best AI. The smarter question: Who will control AI once it enters a real enterprise? This analysis suggests @ServiceNow is making a serious play for that position. Not just with AI language, but with a...
OpenAI Stalls Sora Video AI; Erotic Bot Paused
OpenAI closed its video generator Sora. Erotic chatbot plans on hold “indefinitely” 😅 All sensible decisions… but the headlines are absolute gold.
Testing Siri as Sole Controller for All Agents
lol gonna try using siri exclusively to run all my agents to see how that feels
Ready to Tackle AI Music Grifters on Twitter
Me preparing to go on Twitter and argue with AI music grifters and shills. https://t.co/JABI38OKYD
GenX Reads; Struggle Lies in Comprehension, Not Decoding
It is the comprehension where many struggle. That is what is meant by the latest rage around people saying GeNX cannot read, they can read it is the processing and understanding information in context or understanding the deeper meaning.
OpenAI Resetting Codex Quotas Amid Unfavorable Conditions
Btw simultaneously OpenAI is resetting Codex quotas when the wind blows in the wrong direction.
Claude Praised for Prose; Liberal Journalists Favor Anthropic
50% because Claude is perceived to produce better prose. 50% because journalists skew liberal and Anthropic seems like the more lib-coded of the big AI labs.
Thoughtless AI Makes Prada Campaign Vapid and Uninspired
Prada’s latest AI campaign in collaboration with amazing artist Jordan Wolfson showcases that AI done without much thought behind it always creates a vapid product. I’ve seen some great things created with AI but this is definitely not one of...
Limit Gemini Replies to 15 Words for Clearer Answers
How to tame Google Gemini's logorrhea: constrain your response to 15 words maximum. Makes a world of difference. 😺
AI Poised to Transform Scotland's Public Services
How could #AI change Scotland's public services? by @BBCPhilipSim @bbcnews Learn more: https://t.co/adPsggtzqO #ArtificialIntelligence #MachineLearning #ML https://t.co/igaFFwsiDf
Most People Overestimate AI's Human‑Like Understanding
For those deep into AI, what do you think the biggest misconception about AI is from the general public?
Agentic AI Enables Proactive Inventory Decisions for SMBs
GenAI writes text, but Agentic AI takes action. SMBs need ERPs that suggest inventory pivots before problems hit. Don't just automate: get an agent. 🚀 Schedule a meeting with me: https://t.co/ZsEwnG4p2n #AI #SAPBusinessOne #QuickBooks https://t.co/R7z5eekj1f
Clinical AI API Enables Easy Evidence‑Based Diagnostics
Evidence-based clinical intelligence is becoming a critical infrastructure for many healthcare technology companies. We've made building with our clinical AI agent exceptionally easy with in-app self-serve access to the Glass Developer API. Developers can get started building with our AI today and...
Hospitals Must Ask: Will Clinicians Actually Use AI?
What should hospitals consider before adopting a new AI tool? Start simple: will people actually use it? That’s a key takeaway from our panel at #vMed26: “Translating AI Findings into Real-World Clinical Impact.” https://t.co/dTZgscHIOM
Agentic AI Startups Dominate Berkeley Skydeck Batch 21
Agentic AI startups dominate the roster for Berkeley Skydeck's Batch 21 roster for April 2. https://t.co/4tSv8bKV6M
AI Val Kilmer Demands Its Own Twin for Rendering
AI Val Kilmer's digital twin is now demanding its own digital twin to do the 'hard' rendering. It’s twins all the way down, folks. 👯♂️💻 #AIValKilmer #DigitalTwins #UncannyValleyHigh https://t.co/rgEQPXerPy
Delayed AI Community Digest Launch Next Week
This bums me out. Means far fewer people can really get value from my lists of the AI community. Working on a way to watch the entire AI community here, but getting it up is going slowly. Looks like it...
Comparing Claude Cowork and Microsoft 365 Copilot Use Cases
Anyone using Claude Cowork alongside Microsoft 365 Copilot? Where do you see each one fitting day to day?
We Built the Finance Tool AI Agents Use
Every finance tool is adding an AI chatbot. We went the other direction — we made @tryramp the tool that AI agents use. curl -fsSL https://t.co/dtP4f5DrW5 | bash
AI Errors Highlight, Yet Progress Remains Remarkable
When you look at AI slop, hallucinations, or other misfires, remember that we've had great growth up to the number shown in the left bar. https://t.co/4YeKZIlmFu
Iran Conflict Could Cut AI Supplies by Double Digits
1/ Here’s my baseline assessment of what happens to critical AI components if the Iran war continues. It shows a significant double digit decline. Helium, energy, and a narrow set of fab chokepoints. If the disruption persists, the risk isn’t collapse. It’s a...
AI Simulates Patient Tumors, Transforming Oncology Research
@Ronalfa is speaking at @SynBioBeta in May and you probably already know who he is if you've been paying any attention to the AI x bio space. Ron spent years at @RecursionPharma as SVP of Research and acting CSO, helping build...
AI Adoption Soars, Governance and Risk Lag Behind
AI adoption is accelerating across global organizations, improving productivity, but governance frameworks and risk management systems remain incomplete and inconsistently implemented. https://t.co/RlW0ELL0fT
OpenAI Repo Merges AI Contributions, Codex Stays Anonymous
I explained in today’s @Pragmatic_Eng newsletter: That’s a repo where OpenAI is merging external contributions. Some are made by Claude Code, some with GitHub Copilot, some with Codex. Codex doesn’t add itself as a contributor - on purpose - that’s...
Apple’s Siri Upgrade: Bigger AI Vision, Delayed Rollout
Two things investors should keep in mind as Apple works to do something unique in AI: 1️⃣ The scope of their AI ambitions is bigger than last year’s WWDC comments. It's now expected to include conversational AI and using Siri as...
Claude Code Defaults to Grok 3; Upgrade to 4
FYI for some reason Claude Code will always default you to use Grok 3 and not Grok 4.1 but you should use Grok 4.1 or Grok 4.2 (but that one is more expensive)
Apple's Privacy Focus Limits AI Ambitions, Favors On-Device Models
I’ve argued that Apple is constrained by what it can do with AI by its privacy narrative. Apple *has* trained foundation models, and they have a proprietary ~3BN LLM available on device via Foundation Models Framework. They also offer other...
Tax Revenue From Data Centers Should Cushion AI Job Losses
Americans fear AI-related job losses and resent data centers powering that potential displacement. @MarkWarner says communities should use tax revenue from data centers to soften the blow of this economic transition. https://t.co/F8TsvpEI5n via @techcrunch
AI Benefits Evolve to Year‑Round Support with Guardrails
Contributor Spotlight: LaJoi McClendon — AI-powered benefits moving beyond enrollment to year-round support, with the right guardrails. Read LaJoi's take on p. 26 of our eBook. https://t.co/x6mbQ6zZKz #AI #Talent https://t.co/udEnGmVMon
New AI Platform Uncovers Hidden Cancer Targets, Raises $10M
Most cancer drugs go after the same targets: EGFR. PD-L1. HER2. Not because they’re the best targets. Because they’re the only ones we’ve been able to see. RyboDyn Inc. is going after what’s been invisible. The San Diego team, led by Imad Ajjawi, PhD,...
Top Experts Reveal Keys to Enterprise AI Adoption
Great listen about enterprise AI adoption from two of the smartest minds on the topic
Apple’s Open Siri Boosts Revenue, Justifies AAPL Overweight
Macro: Apple opens Siri to third‑party AIs (iOS27), shifting to an open AI platform and unlocking subscription revenue. Key risks: execution and regulation. Insight: overweight AAPL. — Viktor Kopylov, PhD, CFA More insights: t.me/si14Kopylov
