Today's AI Pulse
-1.png)
Anthropic shuts down Fable 5 and Mythos 5 models under U.S. export controls
Anthropic abruptly disabled its flagship Fable 5 and Mythos 5 models to comply with a U.S. export‑control directive, underscoring the vulnerability of firms that depend on closed‑source AI services. The move triggered a sharp rally in Chinese open‑source offerings as investors shifted to downloadable models that can run on‑premise.
Also developing:
By the numbers: Respond raises $62.5M Series B
PNNL Researchers to Speak About Genesis Mission at National Science Bowl
Pacific Northwest National Laboratory researchers Robert Rallo and Nathan Hodas will address the regional National Science Bowl winners in May, showcasing the Department of Energy’s Genesis Mission that leverages AI and advanced computing for energy innovation and national security. Their talks will illustrate how AI is shifting from a mere tool to a collaborative research partner and explain the inner workings of high‑dimensional AI models. The event, part of a competition that engages more than 10,000 middle‑ and high‑school students nationwide, underscores PNNL’s commitment to STEM outreach.
BMO Turns to AI and Quantum Computing to Predict Earthquakes
Bank of Montreal (BMO) is leveraging artificial intelligence and quantum computing to develop tools that could predict earthquakes and improve disaster response. The bank’s chief AI and quantum officer, Kristin Milchanowski, secured a provisional patent for a quantum algorithm aimed...
Popular Account Turns AI‑generated, Followers Mute It
It's now the 3rd large account on this site, just this week who used to write interesting things and about a year ago I still followed them. Now all their tweets (+threads) are very obviously AI-written, cringe to read (all the...
Atlassian And ServiceNow: The Dominant AI-Enabled IT Management Platforms Lean Into Context Graphs
Atlassian and ServiceNow are cementing their dominance in AI‑enabled IT management by making context graphs the core of their platforms. ServiceNow unveiled a Context Engine that fuses its CMDB, knowledge graph, cyber‑asset graph and Veza access graph, while Atlassian’s Teamwork...
Evaluate AI with Human Collaboration, Not Solo Performance
This is a laudable goal, but it may mean designing, testing, and deploying AI systems differently. For example, why not measure the ability of GPT-5.5 to solve problems *with* a person rather than on its own? (see Stanford's centaur evaluationshttps://digitaleconomy.stanford.edu/project/ai-centaur-benchmarks/)....
Agencies Issue Guidance on Adopting Agentic AI Systems
The Cybersecurity and Infrastructure Security Agency (CISA), the National Security Agency (NSA), and a coalition of international partners have published new guidance on the safe adoption of agentic artificial intelligence systems, especially those built on large‑language models. The document outlines...
Mark Cuban Warns These 5 Job Categories Are at Risk Due to AI
Mark Cuban warns that AI adoption is rapidly endangering five major job categories, especially routine, entry‑level positions. He notes that AI tools now handle data entry, bookkeeping, basic coding, customer inquiries, and simple finance or legal tasks more cheaply than...
AI Should Augment, Not Replace Human Work
the idea that AI is something that by necessity replaces a person in their daily work seems odd to me. the field of AI has of course always sought to understand and replicate human intelligence, but technology is fundamentally about...
AI Spending Inflates GDP without Delivering Real Value
AI and GDP - what does it mean? Sure, AI is “contributing” to GDP — but @davidsacks really nails it below, “Paying people to dig holes and fill them back up increases GDP, as it is currently measured, even though...
How AI Will Change Services
Cooley partner Rachel Proffitt says AI will drive Big Law toward value‑based billing, moving away from traditional billable‑hour models. AI tools can automate routine tasks, improve data‑driven outcome predictions, and enable firms to price services based on results. The shift...
Jensen Warns AI Scare Tactics Hurt Future Talent
Jensen pushes hard on Dario's narrative about AI destroying jobs. I mean, scaring people, and making AI being perceived as a bad thing - is probably not good for AI business? Jensen: "the things that are said are very counterproductive and,...
Speed Turning Tokens Into Value Drives AI Dominance
Google may be the long-term winner of AI because it owns the full stack: chips, cloud, models, data, distribution, and now 1/4 of global AI compute. However, the bigger shift is simple: who can turn tokens into economic value fastest? --...
OpenAI CFO Sees ‘Vertical Wall of Demand’ for Products
OpenAI CFO Sarah Friar told Bloomberg the company is experiencing a "vertical wall of demand" for its AI products, even as it missed some internal revenue milestones and the goal of a billion weekly active ChatGPT users. She emphasized that...
Coatue Has a Plan to Buy up Land for Data Centers, Possibly for Anthropic
Coatue has launched Next Frontier, a venture to acquire land adjacent to large power sources for building AI‑focused data centers, a move that could directly support its stakes in Anthropic and other generative‑AI firms. The initiative already includes a joint...
AI's New Innovator's Dilemma: Products Threaten Their Own Customers
The new innovator’s dilemma A founder told me about meeting with the CEO of a large vertical software company. She walked away realizing he was in a tough place: to fully harness AI in his hero product would mean...
Citi Launches Arc to Scale AI Agents Across the Business
Citigroup has launched Arc, an internal platform that lets developers build and scale AI agents across the firm’s functions and geographies. The platform will first serve internal use cases, with broader employee access slated for later rollout. Arc is part...
Open‑Source Protocol Turns Reactive AI Agents Proactive
A team just open-sourced the missing layer that turns reactive AI agents into proactive ones. It's called World2Agent Protocol (W2A). Think of it as MCP, but for perception. Your agent stops waiting for prompts and starts perceiving the real world like a...
Meta to Acquire Assured Robot Intelligence AI Startup
Mark Zuckerberg and Meta Platforms $META just announced they have agreed to acquire Assured Robot Intelligence "a AI startup developing artificial intelligence models for robots" - Bloomberg
AI, Gene Therapies Drive Market Trends in Eye Care
In 2026 the ophthalmology market is being reshaped by gene‑therapy breakthroughs and the emergence of agentic AI, according to Boston Consulting Group’s Long Sha. Gene‑based treatments are moving beyond rare retinal disorders into chronic conditions such as wet age‑related macular...
BMO's Milchanowski on Quantum's Banking Sector Impact
BMO’s Chief AI and Quantum Officer Kristin Milchanowski told The Close that quantum computing will start reshaping banking within the next few years, with IBM targeting a fault‑tolerant machine by 2029. She highlighted three priority areas—optimization, risk forecasting and cryptography—where...
AI Accelerates Real-World Breakthroughs in Health, Education, Agriculture
I remember growing up reading Sports Illustrated. There was a small column called “This Week’s Sign the Apocalypse Is Upon Us.” With all the dire predictions about AI, it’s important to also spend time recognizing the tremendous pace of innovation &...
Seeking Tenstorrent’s Real‑World Deployments Amid Compute Surge
I’m here in SF to attend the Tenstorrent event. My firm hasn’t researched the company extensively, so I’m here to learn. While previous TT events look like they’re more about architecture and the deep tech, this event name “Deploy” is...
The AI Scaffolding Layer Is Collapsing. LlamaIndex's CEO Explains What Survives.
Jerry Liu, co‑founder and CEO of LlamaIndex, says the traditional scaffolding layer—indexing, query engines, and custom agent loops—used to build LLM applications is collapsing as models become capable of handling massive unstructured data on their own. He argues that this...
Plaud NotePin: Smart Wearable AI Boosts Meeting Notes
I review the Plaud NotePin Wearable AI Note Taker device and find that there's a lot to like with this smart little gizmo. It's particularly helpful for meetings and lectures, as I demonstrate in this detailed video: https://t.co/Q45YtD8M9E #ai #aiWearable...
Join CMU Vision-Language Navigation Challenge at IROS2026
Register for the next CMU Vision-Language-Navigation challenge, which will conclude with a workshop at #IROS2026! Teams will develop systems that allow robots to interpret natural-language instructions and navigate unfamiliar environments. Learn more ⬇️ https://t.co/xXo9RSX12j
Researchers Say This System of 7 Smart Rings Can Translate Sign Language
South Korean researchers have unveiled a wearable system of seven smart rings that captures finger and hand motions to translate American and International Sign Language into text. In tests the prototype recognized 100 distinct signs and could generate sentence‑level translations...
AI's Work Future: From Doom to Optimism, Arguments Explored
Excelllent thread on various views of how AI will affect work, from catastrophic to highly optimistic, and the best arguments for each.
AI-Generated Podcasts Now Number Thousands Each Week
A growing share of podcasts are generated by AI. Thousands every week. Read @ashleyrcarman on the rise of podslop https://t.co/Jp6BNH7CB6
AI Is Forcing Even Insurance’s Most Cautious Players to Move Fast
AI is reshaping the insurance and healthcare sectors at a speed unprecedented for an industry built on caution. Jake Sloan, Appian’s VP of global insurance, highlighted that insurers can now move from pilot projects to full production in weeks, not...
Human Infrastructure Matters as Much as AI Tools
"It is impossible to predict at this point just how AI will transform the workplace. However, one thing is certain: understanding and building the right human infrastructure will be as important as picking the right AI tools." https://t.co/sm48Z9f6lH
AI Mode Shifts Search Toward Trust, Not Clicks
AI Mode transforms search into a trust-driven journey, where clicks validate content rather than initiate discovery. https://t.co/zIgCRWEJat via @gregjarboe, @sejournal
Oscars Tighten AI Rules, Emphasizing Human Authorship
The Academy announced new eligibility rules for the 99th Oscars, mandating that screenplays be entirely human‑authored and that acting awards only recognize performances demonstrably performed by humans with their consent. The academy also reserved the right to request detailed disclosures...
AI Agents Now Automate Every GitHub PR Review
Want AI agents on EVERY GitHub PR? @Macroscope just dropped `Check Run Agents` and it's WILD 🤯 How? Write a markdown file, connect your tools, and let Macroscope enforce your rules, label PRs, and drop inline comments. Time to say goodbye to manual PR...
AI Interfaces Will Replace SaaS Dashboards Soon
Codex is pretty great, and it's an interface battle right now between Codex and Claude/Cowork. Battle for the business/productivity interface keeps heating up. Once these can connect to all your apps/services, there is little reason to visit your SaaS...
GPT-5.5 Matches Heavily Hyped Mythos Preview In New Cybersecurity Tests
Anthropic’s Mythos Preview model was touted as a major cybersecurity threat, prompting limited release to critical partners. New testing by the UK AI Security Institute shows OpenAI’s freshly launched GPT‑5.5 achieved comparable results, scoring 71.4% on expert Capture‑the‑Flag challenges versus...
Former DIA CTO Links AI, Cybersecurity, and Data Center Strategies
We are discussing cybersecurity, data centers and AI with the former CTO of the Defense Intelligence Agency https://t.co/vDEsjpoKep
Academy Mandates Human-Authored Screenplays Amid AI Rules
The Academy further addressed generative AI more broadly, reserving the right to request additional information...In the writing categories, the rules now formally codify that screenplays must be human-authored to qualify. https://t.co/mO1YgIzPYl via @variety
Apply, Lie, Move In: AI Is Making Rent Fraud Easier
Landlords are facing a surge in AI‑generated rental fraud, with scammers easily producing fake pay stubs, IDs and bank statements. The FBI estimated online real‑estate fraud cost about $275 million in 2025, while each fraudulent application costs roughly $15 K to resolve....
Choose the Best AI Code Tool, Not Polls
you know what all of these "which is better" polls are silly use codex or claude code, whatever works best for you i am grateful we live in a time with such amazing tools, and grateful there is a choice
AI Reveals Need to Align Skills with Cognitive Well‑Being
AI is exposing a gap between what organizations expect and what people can sustain over time. Closing it means linking skills with cognitive well-being, otherwise pressure builds and execution weakens. Source @McKinsey Link https://t.co/mUYF7giLBR via @antgrasso https://t.co/8aw7ZgjBIu
DARPA Issues RFI on Embedding Intelligence Into Robotic Materials
DARPA has issued a Request for Information seeking concepts for materials that embed sensing, computation, and actuation directly into robotic structures. The agency aims to shift intelligence from centralized processors to the hardware itself, enabling faster, lower‑power, and more resilient...
Embodied AI Enables Fully Autonomous Surgical Procedures
excited about this. embodied AI unlocks the ability to perform surgery completely autonomously. @r_mcguire @dylanmcguir3 are absolute maniacs. proud to be backing them early. the future of surgery is bright.
News Industry At Odds With Free Speech Advocates Over Amazon's Perplexity Ban
A federal judge granted Amazon a preliminary injunction blocking Perplexity's AI‑driven shopping agent from accessing Amazon accounts, citing likely violations of the Computer Fraud and Abuse Act. Perplexity is appealing the order to the 9th Circuit, while free‑speech groups argue...
AI Constraints Must Come Before Deployment, Not After
Anthropic announced its most powerful model, Claude Mythos, but chose not to release it because the safety‑constraint infrastructure needed to control its outputs does not yet exist. Testing revealed the model could uncover thousands of zero‑day vulnerabilities across major operating...
OpenAI Codex Master Guide: How to Build Apps With AI
OpenAI’s Codex, now powered by the GPT‑5.5 model, is positioned as a full‑stack coding agent rather than a simple code‑completion tool. It can read project files, edit code, run commands, execute tests, and return diffs through integrations in a visual...
XAI Launches Grok 4.3 at an Aggressively Low Price and a New, Fast, Powerful Voice Cloning Suite
Elon Musk’s xAI unveiled Grok 4.3, its latest proprietary large‑language model, alongside a new Custom Voices voice‑cloning suite. Grok 4.3 introduces a permanent reasoning mode, a 1 million‑token context window, and aggressive API pricing of $1.25 per million input tokens and $2.50 per...
AI Godfather Geoffrey Hinton Says We Must Convince AI that It’s Our Mother
Geoffrey Hinton, the AI pioneer who left Google in 2023, warned that super‑intelligent AI could pose an existential threat within the next two decades. At the DiscoveryX conference in Toronto, he suggested embedding "maternal instincts" in AI so it would...
PNNL Scientists Leverage AI to Optimize Glass Formulas for Liquid Radioactive Waste
Scientists at Pacific Northwest National Laboratory have applied AI-driven machine learning with active learning to design optimized glass formulations for immobilizing Hanford’s liquid radioactive waste. The new models increase waste loading, enabling roughly 1% more waste per 20% loading increase...
How Agentic AI and New Tax and Audit Research Tools Will Shape 2026
Agentic AI is set to overhaul tax and audit research by 2026, moving beyond simple search to systematic reasoning and workflow automation. The new generation of tools, exemplified by Thomson Reuters' CoCounsel Tax and CoCounsel Audit, embed authoritative content, contextual...
Good Conversation, Smart Music … Oh, and Directions. (GM)
Google is rolling out its Gemini audio chat interface to replace Google Assistant in GM vehicles built from 2022 onward. The AI‑powered system enables natural, two‑way conversation for navigation, music, podcasts, translation and casual chat, aiming to keep drivers’ hands...