Know What's Happening in AI
Know What's Happening in AI
AI‑powered digital employees are now handling everyday office tasks
AI‑driven digital workers are already managing routine duties such as inbox summarization, calendar coordination and expense reporting. An EY survey finds 84% of employees are open to using agentic AI, while 51% worry it could make their jobs obsolete. Analysts forecast 1.3 billion intelligent robots in the global workforce by 2035.
Also developing:
By the numbers: OpenAI raises $12B funding round
Cohesity, together with ServiceNow and Datadog, is launching a recoverability service that can detect and roll back damage caused by agentic AI in enterprise environments. The solution leverages immutable snapshots and API‑driven restorations to return files, databases, vector stores, and AI model states to a trusted point‑in‑time. Competitors such as Rubrik and Cisco already offer similar rollback capabilities, signaling a nascent market for AI‑focused disaster recovery. Gartner predicts 40% of enterprise applications will embed task‑specific agents by 2026, underscoring the growing need for such safety nets.
Local colleges and universities across the U.S. are rapidly building AI capabilities, yet their expertise remains largely untapped by city governments and nonprofits. While higher‑education institutions are governing AI internally and creating a pipeline of AI‑trained talent, public sector entities...
Singapore is building a next‑generation, AI‑enabled Eastern General Hospital, scheduled to open around 2029. The smart hospital roadmap integrates digital platforms, AI, robotics, wearables, and real‑time location systems to enhance clinical and operational efficiency. Telehealth and generative AI tools aim...
Oracle’s leadership used its earnings call to dismiss the so‑called SaaSpocalypse, arguing that AI‑driven disruption will spare the company. Larry Ellison said Oracle’s AI‑enabled coding tools let it build end‑to‑end, agent‑based solutions for sectors like healthcare and finance. CEO Mike...
Magister sneak peek #2: Skill catalog One-click install for agent skills. No terminal, no config files, no copy-pasting commands. Browse the catalog, hit install, done. Your agent immediately gets new capabilities — web search, browser control, frontend design, long-term memory, workflow automation,...
The 2025 Cost of a Data Breach Report shows a 9% drop to $4.44 million, driven largely by security AI and automation that slash detection times. Yet organizations that fully automate see breach costs up to $1.9 million lower than manual peers,...
Google DeepMind unveiled its latest generative AI model, Genie 3, at the Game Developers Conference in San Francisco. The presentation, “The Future of Playable Worlds With Google DeepMind,” filled the Moscone West room to standing‑room‑only capacity, with more than 100 developers turned...
In the 1950s, you could simultaneously think A) nuclear energy was good, and B) we were in great danger of a nuclear war. In the 2020s, you can simultaneously think A) AI agents will be a net good for the...
Cisco’s 2026 State of Industrial AI report, based on 350+ manufacturing leaders across 19 countries, reveals that infrastructure shortfalls are throttling AI rollout in factories. While 96% view reliable wireless connectivity as essential, 56% cite frequent outages that disrupt operations....
Singapore‑based Princeton Digital Group, backed by Warburg Pincus, announced plans to raise up to $5 billion in debt this year. The proceeds will finance a multi‑country AI data‑center build‑out in India, Indonesia, Malaysia and Japan. PDG already operates AI‑focused facilities in seven...
Professional services firms are confronting AI‑driven efficiency that compresses delivery time, prompting a shift from hourly billing to outcome‑focused pricing. INSIGHT proposes a five‑level pricing maturity framework that guides firms from role‑based rate cards to fully value‑based models. The roadmap...
At MWC Barcelona 2026, Huawei announced a shift from AI experimentation to large‑scale operational use in enterprises. It showcased 115 industrial intelligence demonstrations and introduced 22 new industry solutions across energy, manufacturing, finance, transport and retail. Central to the rollout...
U.S. Indo‑Pacific Command (INDOPACOM) built its war‑gaming and command‑control workflows around Anthropic’s Claude model, only to lose access after a Trump‑issued ban on Anthropic tools for federal agencies. The ban sparked a lawsuit from Anthropic alleging illegal retaliation against the...
A developer used AI to generate Lightning Web Components, achieving near‑perfect code for functional requirements. When asked to redesign the UI for tablet use, the AI repeatedly produced unusable layouts despite multiple prompts. Even feeding AI‑generated design advice into another...
Ford Motor announced the launch of Ford Pro AI, an artificial‑intelligence platform for its Pro commercial‑vehicle subscribers. The system will analyze more than one billion data points each day, covering vehicle health, driver behavior, route efficiency and fuel use. It targets...
Ford unveiled Ford Pro AI, a generative‑AI chatbot embedded in its Pro telematics platform, to turn fleet telemetry into actionable insights for its 840,000 paid subscribers. The model‑agnostic service runs on Google Cloud, operates in read‑only mode and still requires human oversight,...
Salesforce unveiled new Agentforce capabilities, including Agentforce for Communications—a suite of pre‑built AI agents tailored for telecom sales and customer‑service workflows. It also introduced the Agentic Work Unit (AWU) metric, which counts each discrete AI‑driven task, reporting 2.4 billion AWUs to...
The article warns that AI is currently perched on the Gartner Hype Cycle’s Peak of Inflated Expectations, where lofty promises outpace real capabilities. It highlights how over‑hyped features—especially language‑specific nuances—often miss the mark, likening them to the short‑lived teardrop phones...
Wikipedia is included in virtually every major LLM training dataset and I built something to help you set it up. When ChatGPT, Claude, Gemini, or Perplexity answer "Who is an expert in ___?" — Wikipedia and Wikidata are baked into how...
Silicon Valley’s newest wave of health‑tech startups is targeting America’s $5 trillion, fragmented medical system. Patients increasingly skip or delay care because of cost, with roughly one‑third reporting financial barriers. These disrupters are leveraging artificial intelligence to streamline diagnosis, pricing transparency,...
The FDA’s artificial‑intelligence advisor announced a shift toward evaluating AI‑driven health technologies on a case‑by‑case basis, focusing on each tool’s specific function rather than applying a one‑size‑fits‑all rule set. This risk‑based approach will tailor regulatory requirements to the actual clinical...
I’m starting to experiment with Agent Teams for software development. One writing code. One reviewing. One testing. Feels a bit chaotic right now but incredibly powerful. Really interested to hear about how this is working for larger enterprises. Hundreds of team members each...
Silicon Valley is adding AI inference compute as a fourth element of compensation, alongside salary, bonus and equity. Candidates now ask about dedicated GPU or token budgets for running models like Codex, reflecting growing scarcity of inference resources. Companies such...
Global health systems have reached roughly 65% of the key AI readiness indicators, according to HIMSS chief scientific research officer Anne Snowdon. Governance frameworks and workforce development programs have improved, yet substantial gaps remain in data infrastructure, interoperability, and change‑management...
The Aerospace Industries Association released a white paper positioning agentic AI as a powerful enabler for faster defense procurement and supply‑chain decisions. Tim White emphasizes that AI will not replace culture, processes, or workforce but can augment them when paired...
The paper introduces CONSTORY‑CHECKER, an automated pipeline, and ConStory‑Bench, a 2,000‑prompt benchmark, to evaluate narrative consistency in long‑form story generation by LLMs. The four‑stage system extracts suspect spans, pairs conflicting statements, generates evidence chains, and produces anchored reports. Evaluation across...
Gracenote, the leading entertainment metadata provider, has filed a federal lawsuit against OpenAI, alleging that ChatGPT was trained on its proprietary data without permission. The complaint asserts that OpenAI reproduced Gracenote’s detailed descriptions and the relational framework that organizes its...
I love that clanker for LLMs and coding agents is catching on. I always imagine some old rusty robot hacking away on one of those beige gray keyboards in the darkness now.
Female entrepreneurs are increasingly driving AI innovation despite receiving a fraction of venture capital compared to male peers. The Inc. Female Founders 500 list spotlights 23 women leading breakthroughs across sectors such as enterprise software, health care, digital fashion, and scientific...
The so‑called “SaaS‑pocalypse” erased over $1 trillion in market value, dragging Xero’s shares down 23% YTD, yet CFO Claire Bramley says the New Zealand‑based accounting SaaS remains fundamentally strong. Xero posted 20% year‑over‑year operating‑revenue growth and a 10% rise in global subscriptions,...
Hasbro CEO Chris Cocks told The Verge that artificial intelligence will never be used to design Magic: The Gathering or Dungeons & Dragons, even though the company employs AI for other licensed toys such as Peppa Pig and Optimus Prime. He highlighted...
Proud to share that my first-author paper, “Optimizing Spectrum Allocation Using Deep Reinforcement Learning in Heterogeneous Wireless Networks,” has been published in Internet Technology Letters, a Wiley journal focused on emerging internet and communication technologies. This work explores how deep...
🗞️@AnthropicAI is expanding to Australia 🇦🇺and New Zealand🇳🇿, opening a Sydney office in the coming weeks—its fourth APAC office alongside Tokyo🇯🇵, Bengaluru🇮🇳, and Seoul🇰🇷—to support growing demand and better serve the region’s AI ecosystem. https://t.co/DGe5jbueVm
Salesforce unveiled the Agentic Work Unit (AWU) on its Feb 25, 2026 earnings call, defining it as a single AI‑driven task that converts raw intelligence into actionable work. The company reported processing over 19 trillion AI tokens into 2.4 billion AWUs, with 771 million delivered...
The best AI strategy? One your users don't even notice. If you want immediate ROI, "slipstream" AI behind the scenes. If they have to change their workflow to find the value, you've already lost them. 🤫 Details here: https://t.co/KP9ClquPul #CIO #AI #Productivity...
.@rollingstone remembers babyagi 🥰 (in an article questioning the role of machines vs humans from yesterday) https://t.co/H71nwLSBCv
Researchers at the Arkansas Agricultural Experiment Station unveiled ChicGrasp, a dual‑jaw robotic gripper that learns to grasp, lift, and hang chicken carcasses by imitating human motions. The system employs a diffusion‑policy imitation‑learning algorithm and camera perception, achieving an 81% success...
We're having the Notion AI team (including at long last @simonlast) on the pod Thursday. send me all your questions on this + Notion AI! not an ad, just a fan. Notion is probably the most impt knowledge work agent lab in...
AI coding assistants may influence which languages developers use https://t.co/NUSmA6RF0d < seems possible, even likely. If you're polyglot, do you find that you favor one specific language for AI-generated code?
In this episode Adam covers a whirlwind of tech news, from the shocking AWS data center attacks to the release of OpenAI's GPT‑5.4, which Augment Code praises for its agent‑workflow capabilities. He highlights a demo where a brain‑computer interface runs...
Free professional advice: If your email is longer than 100 words, paste the draft into Claude, ask it to reduce the word count by 50% without abridging substantive content, review it, and then send it.
UK government issues open call for experts to shape AI energy policy #energysky -- via pv magazine global: https://t.co/hQ1EUBbdg9
“Some orgs. are deploying AI w/out a clear sense of the benefits it will bring—b/c they’re driven by herd mentality. They may end up w/ employees who are more productive yet their organizational contribution is unchanged.” 🔗 https://t.co/5j1dKP8sIs #leadership #management #HR https://t.co/48qplP9WO3
One way to think about this is: Agents are machine tools for knowledge work, and artisans are now shifting to being machinists, just like in factories a century ago.
Meet SIMA 2 — The Ultimate #AI Explorer for Virtual #3D Worlds by @GoogleDeepMind #ArtificialIntelligence #MachineLearning #ML #Technology https://t.co/5SkSVsD8NG
METR needs a benchmark on how effectively AI agents can DESTROY THE INTERNET, and labs need to optimize to hit that benchmark
One reason I'm long-term optimistic on alignment is that I don't think that intelligence and human morality are actually as separable as econ models (and AI doomers) tend to assume. I think this is an early sign of that. (In the...
Says the man who claimed for years to care about AI risk and now runs the most reckless of the “frontier model” companies. I just wrote a lengthy critique of Amodei in my newsletter, but this coming from Elon is...
Patreon CEO Sounds Off on AI: ‘I’m Both Amazed and Furious’: As a Creator, ‘I’m Angry That We Aren’t Being Paid’ for Value of Contributing to AI Models https://t.co/rYw6dV5WNW via @variety
Noetic (@getnoetic) uses AI agents to automate end-to-end compliance workflows, so hardware products can get to markets 10x faster. Congrats on the launch, @togao0, @henrypzheng & @J4ckJ1Y! https://t.co/yDKh3UF6ks https://t.co/bMZ4rFcPBK
