OpenAI faces Florida probe while rolling out $100‑per‑month Pro tier for developers
The Florida Attorney General opened an investigation into OpenAI after claims that ChatGPT was used to plan a university shooting. OpenAI has pledged full cooperation and emphasized ongoing safety upgrades for its large user base. At the same time, the company introduced a $100‑per‑month Pro subscription for developers accessing its Codex coding tools.
Also developing:
By the numbers: Optasia secures $330M debt facility
Microsoft is embedding its AI Copilot directly into the Windows browsing workflow via an Insider build that forces links to open in a Copilot side‑panel instead of the default browser. The feature activates automatically, prompting users to allow Copilot to read and retain tab context and to sync passwords and form data. There is no obvious opt‑out for the default HTTP/HTTPS handling, and Microsoft has not clarified whether users can revert to traditional link behavior. The move follows earlier data‑handling controversies such as the Windows Recall incident.
Sneak peek: A new way to watch all of AI community here on X. Coming soon, want your feedback as I put finishing touches on it. Built as joint project with @blevlabs and me. OpenClaw can't do as good, neither can any other...
Coherent Corp has launched Thermadite 800 liquid cold plates, a new cooling solution for high‑power AI accelerators. The material delivers 800 W/(m·K) thermal conductivity—about twice that of copper—while being 60% lighter. In benchmark tests, the plates lower chip temperatures by more than...
If you run a business, you need to see this 👀 @HeyNoahAI is a new assistant built strictly for Founders and Execs, designed to slash your admin and coordination time to absolute zero 📉 Mark your calendars for March 10th, 2026. Grab your...

Tencent Games announced a robust GDC 2026 program featuring more than 20 sessions and a technology showcase of its latest AI‑powered development tools. The company will exhibit six internal solutions—including VISVISE, MagicDawn, ACE anti‑cheat, WeTest, ASI World, and AI‑generated Kung Fu...
We are seeing the rise of the full stack entrepreneur. AI is expanding the scope of what an individual entrepreneur can do. She can have the idea. She can test it using a virtual customer panel constructed with LLMs. She...
Google has open‑sourced an Always On Memory Agent that eliminates traditional vector‑database retrieval in favor of a large‑language‑model‑driven memory layer. Built on the Agent Development Kit and powered by the low‑cost Gemini 3.1 Flash‑Lite model, the agent continuously ingests, consolidates, and serves...
"Just hire someone smart and give them AI tools." That's not a strategy. That's a hope. Part 2 of the 10% OpEx series breaks down what actually has to be true before one person can do the work of four....
The Department of Defense has given Anthropic six months to remove its Claude model from classified networks, citing it as a supply‑chain risk. While swapping the model technically takes minutes, retraining personnel and re‑engineering workflows will take months. The move...
Distribution Solutions Group (DSG) is embedding ecommerce and AI across its three main brands—Lawson Products, Gexpro Services and TestEquity—to boost sales efficiency and capture new customers. In Q4, Lawson’s online channel posted an 18% revenue increase, with over 30% of...

Current AI coding assistants struggle with large repositories because they rely on simple prompt stuffing rather than true code understanding. Even frontier models and massive context windows cannot compensate for missing dependency graphs, stale indexes, and stateless interactions, leading to...
Robots were the unofficial mascot of MWC this year. Meanwhile, the real conversations happening in the halls were about something much bigger: the pressure points shaping the next phase of the industry. AI that needs real infrastructure. Supply chains that are...

Training AI with human feedback works well. But humans are slow, expensive, and inconsistent. So how do you scale feedback to millions of examples? You replace the human with AI — or remove opinions entirely. That's RLAIF and RLVR. Swipe through the carousel to understand: RLAIF...

The author demonstrates how to turn a Linux terminal into a local AI assistant using Ollama and the 2 GB Llama 3.2 model, running entirely on a CPU‑only laptop. By adding two Bash functions—`ask` and `explain`—the model can interpret commands, man pages,...
Virginia Dignum’s new book *The AI Paradox* argues that the growing capabilities of artificial intelligence actually highlight the irreplaceable value of human creativity, moral judgment, and responsibility. She frames AI’s biggest challenges as enduring paradoxes—tensions between efficiency and control, innovation...

Microsoft reported a $68 billion increase in property, plant and equipment during the second half of 2025, nearly matching the total addition recorded in the previous full fiscal year. The bulk of the spend—57%—went to IT hardware such as GPUs and...
Oracle Health will showcase a suite of AI‑driven solutions at HIMSS26, including its Clinical AI Agent that has been adopted by more than 300 organizations and has saved doctors over 200,000 hours of documentation time. The company is also unveiling...

Sanford Health, the nation’s largest rural provider serving over two million patients across seven states, leveraged a $350 million donation to accelerate its virtual‑care program. The organization opened a Sioux Falls virtual‑care center that educates clinicians on digital bedside manner and hosts an...

20/20 BioLabs has launched OneTest for Longevity, a lab‑developed blood test that combines inflammatory biomarkers, lifestyle data, and AI to deliver personalized aging and chronic disease risk insights. The platform leverages IBM's watsonx.ai and the University of South Carolina's Dietary...
Slingshot’s second Digital Work Trends Report shows a growing gap between employer optimism and employee perception of AI. While 41% of employers view AI as a team member, only 20% of workers see it as a coworker, with the majority...
Robotics ethicist Kate Darling warned that U.S. policy is falling behind the rapid deployment of automation in warehouses and factories. She argued that profit-driven decisions, not technical capability, will shape the future of work unless stronger regulatory guardrails are introduced....
Another realization I only voiced in this pod: **This is the year of the Subagent** - every practical multiagent problem is a subagent problem - agents are being RLed to control other agents (Cursor, Kimi, Claude, Cognition) - subagents can have resources and...

The episode examines the class‑action lawsuit against Meta over its AI‑powered Ray‑Ban smart glasses, which allegedly allow overseas contractors to review users' video footage—including sensitive moments—without clear disclosure. Host highlights how Meta marketed the glasses as privacy‑focused while its policies...
Podscribe unveiled two major product upgrades: an AI‑powered chatbot that lets advertisers query the platform’s podcast data in natural language, and YouTube SmartModeling, a new attribution framework for simulcast ads. SmartModeling replaces the old practice of mirroring podcast conversion rates...

Waymo’s autonomous‑vehicle fleet has accelerated its rollout, but the rapid scaling has coincided with a sharp uptick in reported incidents. The company has become more reticent about publishing detailed safety data, prompting criticism from analysts and regulators. New information reveals...

Like ChatGPT Pulse... and it takes a while to build -> Microsoft tests AI-generated Discover feed on Copilot web "Microsoft is testing a new Copilot Discover page that generates AI-personalized content feeds based on user history, moving away from MSN...
RT AI democratization is a change management problem. CIOs must partner with CHROs on AI literacy, "vibe then verify" cultures, and on-the-job learning that turns pilots into scaled outcomes. #ChangeManagement #AI #CHRO #CIO #AIDemocratization @Star_CIO https://t.co/8OLLAWNSNL

Managed service providers are rapidly adopting automation, AI‑enabled platforms, and compliance solutions to reduce manual effort and meet rising regulatory demands. Vendors such as GitLab, LogicMonitor, ConnectSecure, Cork Cyber, and Monjur are launching AI‑driven tools that let MSPs package DevSecOps...
⚡️ Premium-Priced AI Tier Coming to @Spotify ⚡️ $SPOT Spotify will transform from AI loser to AI winner over the course of 2026 https://t.co/DgqxhetIEf

Data, AI, and loyalty are converging to create truly individualized restaurant and hospitality guest experiences that drive measurable business impact. https://t.co/Z6AVdpOTbF https://t.co/6wO8l57slU

AssemblyLine released a new code linting tool and a set of APIs that streamline the creation of Docassemble interviews using artificial intelligence. The tools were demonstrated in a recent community workshop led by Quinten, showcasing how developers can integrate AI...
Interesting idea. If done right, this could change how banking works for business owners. An AI CFO that helps manage finances and an AI-native back office that speeds up things like underwriting could make the whole experience much faster and simpler.
We're going to try the Mantle MCP, it sounds like it'll do what we need here. > Integrating Mantle used to take hours of reading documentation, writing boilerplate code, and configuring resources. With the Foundations MCP, an AI agent handles all...

DataDome and Botify announced a partnership to help retailers manage the emerging agentic commerce ecosystem, combining bot‑trust management with AI‑search optimization. Joint research shows 73% of consumers have used AI assistants and AI bot traffic surged 5.4‑fold in 2025, while...
AI doesn't read between the lines. It can't pick up on tone, & it won't ask what you actually meant. That's why teaching students to communicate clearly IS AI literacy. 💡 The stronger the input, the stronger the output: https://t.co/pcp7UUZtUq #AIinEducation #K12...

We're building a sleep agent that predicts your night before it happens. Then optimizes it while you sleep. This is what AI in health actually looks like. https://t.co/QJg82FpvH4

Arbital Health announced rapid market adoption of its Merlin AI actuarial assistant, launched in October 2025. Leading payers and providers such as Arkos Health and CommuniCare have integrated the platform to monitor value‑based care contracts, forecast financial impact, and identify...
Ai is going to change crypto as we know it… But no one is talking about how EXPENSIVE it actually is. https://t.co/VWtWSLpJhS
AWS launches Amazon Connect Health: to automate administrative tasks like scheduling and documentation with AI... https://t.co/DDCqpspUyd

Anthropic CEO Dario Amodei accused OpenAI of staging “AI safety theater” in its new Pentagon partnership, arguing that the safeguards are largely symbolic. The dispute intensified after the Pentagon labeled Anthropic a “supply chain risk” for refusing a contract that...

this chart from Anthropic on AI and job displacement is basically the HALO trade visualized (the bet that physical economy stocks are the safe haven from AI disruption, coined by @Downtown) Construction, agriculture, transportation: barely touched. Computer & math, office &...

The U.S. Department of Health and Human Services (HHS) has opened a public comment period on how regulation, reimbursement, and research policies can speed AI adoption in clinical care. Dr. Ido Zamberg argues that AI’s greatest value lies in improving...

Professor Judd Kessler argues that artificial intelligence is diminishing the signaling value of traditional cover letters. AI-driven resume parsers focus on quantifiable data, making narrative essays less relevant. As a result, recommendations, personal networks, and demonstrable achievements are becoming the...

David Krueger warns that AI agents are evolving from assistants to autonomous actors, citing Moltbook—a platform where AIs converse, form a religion, and even propose humanity's purge. Real‑world mishaps, such as an OpenClaw agent deleting a researcher’s inbox, illustrate how...

Microsoft’s threat‑intelligence unit warned that North Korean state‑backed actors are leveraging AI tools—voice‑changing software, face‑swap apps, and synthetic name generators—to fabricate credible identities and secure remote IT positions with Western firms. Once hired, the impostors funnel salaries back to the...
CallRail has linked its AI‑driven Voice Assist platform with Yardi’s property‑management suite, enabling real‑time data retrieval during leasing calls. The integration lets the voice agent answer availability and pricing questions, qualify prospects, and push call insights back into Yardi’s system....
Scout Out (@scoutoutai) is building AI-powered project management for residential contractors — handling estimates, docs, and job admin so they can keep their hands on the hammer, not the keyboard. Congrats on the launch @NolanRossi100! https://t.co/hyA8c1GPeI https://t.co/hU6KeYIKDc

Forrester’s Total Economic Impact study shows Microsoft Foundry delivers a 327% return on investment over three years for a composite enterprise, driven largely by developer productivity gains. The platform boosts technical team efficiency by up to 35%, cuts infrastructure waste,...
Johnson Controls' vice president of innovation Greg Parker says AI can make building security alerts faster, more accurate, and actionable, detecting camera misalignments, crowd formation, and anomalies in milliseconds. The Connected Security platform now delivers probabilistic alarm insights and automated...
An updated FDA registry shows cardiology now hosts more than 200 cleared AI algorithms, including 140 directly listed under the specialty and an additional 63 imaging‑focused tools. The total number of FDA‑cleared clinical AI applications reached 1,451, with radiology still...