Today's AI Pulse
Odyssey lands $310M Series B, valued at $1.45B
Odyssey announced a $310 million Series B round that values the AI lab at $1.45 billion. The round was led by Natural Capital and included backers such as Amazon, AMD Ventures, GV, EQT and IQT. In parallel, Odyssey struck a preferred‑cloud deal with AWS to use Trainium silicon for large‑scale model training.
Also developing:
Can a Fake Brand Win in AI Search? New Experiment Says Yes
SE Ranking’s 16‑month experiment introduced a fictitious brand across 20 sites to gauge how large language models surface new content. Within the first month, AI systems generated 15,835 answers, with 96% of visibility tied to branded queries and deep‑guide, review, and comparison pages earning the most citations. Perplexity indexed fresh pages in 1‑3 days but often cited supporting domains, while Google AI Mode consistently placed the brand at #1 for 90% of branded prompts. The findings reveal that early brand narrative can be shaped in AI search, but broader topic authority remains dominated by established sites.
Bloomberg Opinion’s Dave Lee Says an OpenAI Bubble Is Not an AI Bubble After Wall Street Reaction to Growth Report
Bloomberg Opinion columnist Dave Lee argues that Wall Street’s sharp sell‑off in AI stocks after a WSJ report on OpenAI missing growth targets is misreading the broader AI market. He notes ChatGPT’s U.S. market share fell from 55.4% to 38.3%...
PTC Launches Windchill AI Assistant for Natural Language Search
PTC announced the Windchill AI Assistant, a generative‑AI chat plugin for its Windchill product‑lifecycle‑management platform. The tool lets engineers and manufacturers ask natural‑language questions and receive contextual answers drawn directly from PLM documents, complete with source citations and governed by...
Cleveland-Cliffs Partners with Palantir on AI-Driven Steelmaking
Cleveland‑Cliffs announced a three‑year partnership with Palantir Technologies to embed its AI platform into the company’s flat‑rolled steel production, planning, and order‑entry workflows. The AI integration aims to provide real‑time data coordination across the firm’s 32 North American sites, boosting...
LG Electronics and Nvidia Are in Talks on Robotics, AI Data Centres, and Mobility
LG Electronics confirmed exploratory talks with Nvidia to collaborate on robotics, AI‑powered data‑center cooling, and automotive mobility solutions. The discussions follow Nvidia senior director Madison Huang’s visit to LG’s Seoul headquarters and align with both firms’ strategic pushes into physical‑AI....
AI Will Gift Us More Time and Longevity
The Time of Your Life: How AI Will Give You More of It | Michael Nuschke | TEDxAjijic "What if you won the "Time" lottery? Guess what...you just did. Artificial Intelligence (AI) will transform our lives, taking over the mundane tasks...
Enterprises Plan Rapid Growth for AI Factories and AI at the Edge, Survey Finds
A Deloitte survey of 515 senior IT and business leaders shows enterprises plan to double AI factories and edge deployments by 2028, with 73% targeting at‑scale AI factories and 72% aiming for edge AI. Monthly AI token consumption is expected...
AI's Magic Unveiled: Understand Limits, Bias, Hallucinations
🤖 AI can feel almost magical, but to use it well, we need to understand what it is, how it works, and where its limits really are. In this video, I explain AI in simple terms, including how tools like...
Accountex London Launches AI Hub in Partnership with Hello AI Collective, Powered by Lenovo
Accountex London, the world’s largest accounting and finance expo, is launching an AI Advisory Hub at Stand 1650 on May 13‑14, 2026. The hub, created with Hello AI Collective and powered by Lenovo’s Solutions and Services Group, offers free, drop‑in AI guidance...
The DMA’s AI Dilemma: Too Soon, Too Late, or Both?
The European Commission’s first review of the Digital Markets Act (DMA) concludes the law remains fit for purpose, but warns it could age rapidly as AI reshapes digital intermediation. AI embedded in existing gatekeepers can be regulated under current DMA...
AI Agents May Self‑police as Humans Outsource Judgment
There was an article this week in the Wall Street Journal about how people are outsourcing their judgment to AI. While I agree that this is probably happening, I wonder whether people will have so many AI agents that counterbalance...
UK Government Should Clarify Whether It’s Investing in AI Companies that Use Unlicensed Training Data, Say Campaigners
The UK’s new Sovereign AI Unit, backed by a £500 million (~$635 million) fund, is investing in seven AI start‑ups and courting 30 more. Campaigners, led by Lords’ peer Beeban Kidron, warn the unit has not clarified whether funded firms must license...
Orca AI and SHI Team up on Next-Gen Autonomous Vessel Tech
Orca AI and Samsung Heavy Industries (SHI) have signed a Memorandum of Understanding to fuse SHI’s autonomous ship hardware with Orca AI’s AI‑driven maritime platform. The collaboration will embed SHI’s SVISION berthing assistance and autonomous speed control into Orca’s SeaPod...
The Agentic AI Advantage with Informatica
A recent DBTA webinar highlighted that while 85% of organizations have deployed AI agents, most encounter a productivity paradox due to constant human coordination and inadequate data governance. Speakers from Informatica emphasized that 65% of firms lack AI‑ready data or...
Actian Debuts VectorAI DB, Giving Developers Data Ownership Across Edge, On-Premises, and Cloud Environments
Actian, the data and AI arm of HCLSoftware, has launched VectorAI DB, a portable vector database designed for production AI in regulated, disconnected, and edge environments. The solution delivers low‑latency vector search on embedded systems, edge devices such as NVIDIA...
OpenAI Forecasts $8 ChatGPT Go Tier Will Drive Consumer Subscribers to 122M This Year as Plus Subscribers Fall 80%
OpenAI projects its $8‑per‑month ChatGPT Go tier to attract roughly 112 million subscribers this year, inflating total paying users to 122 million. In contrast, the $20 ChatGPT Plus plan is forecast to tumble 80% to about 9 million subscribers. By year‑end, Go is expected to...
Anthropic More than Doubles Its Public Estimate of Average Daily Claude Code Spending From $6 to $13 per Developer
Anthropic has revised its Claude Code pricing guidance, estimating that an average enterprise developer now spends about $13 per active day on token usage, up from the previous $6 estimate. The new figures show 90% of users staying below $30...
Master Time‑Series Forecasting in Just 3 Minutes
90% of data scientists are not confident in time series forecasting. I should know. It took me 3 years. Let me crush your confusion in 3 minutes.
GitHub Copilot’s Price Spikes Expose Lack of Own Coding Model
Microsoft / GH Copilot is now paying the price for not owning a foundational model of their own that is good at coding. They now need to resell the markup of OpenAI, Anthropic, Google Cursor was v smart to build...
OpenAI Researchers Explain Why Math Is the Road to AGI
OpenAI researchers Sebastian Bubeck and Ernest Ryu say mathematics has become the litmus test for artificial general intelligence, noting that LLMs have leapt from elementary arithmetic to solving olympiad‑level and research problems in just two years. Ryu used ChatGPT to...
7 Ways Codex Is Redefining Knowledge Work
Codex is a super‑app that connects local and cloud files directly to an AI, allowing users to query entire knowledge bases without manual copy‑pasting. Its Persistent Memory feature remembers client contexts, slashing project setup time from twenty hours to three...
The AI Super-Cycle
The post declares that we are at the outset of a 20‑year AI‑driven capital‑expenditure super‑cycle that will fundamentally reshape how work, value and defensibility are defined across the global economy. It introduces a 25‑layer technology investment stack—from services and physical...
AI Falls Short on Differential Diagnosis, Despite High Accuracy Rates
A JAMA Network Open study from Mass General Brigham evaluated 21 large language models on structured patient scenarios. The models achieved over 90 % accuracy in identifying the final diagnosis when given complete clinical information, but they struggled to generate appropriate...
SAS Finds 500+ Leaders Eye Quantum AI for ROI
SAS’s 2026 survey of over 500 global executives shows uncertainty about practical quantum AI applications now outweighs cost concerns, marking a shift from earlier affordability focus. The study ranks real‑world use‑case ambiguity as the top adoption barrier, followed by cost,...
AI and Investing: David Trainer’s Presentation at Harvard Business School
Assistant Professor Yuan Zou hosted a Harvard Business School panel featuring New Constructs CEO David Trainer, Hudson Labs' Kris Bennatti, and Alumni Ventures' Ron Levin to discuss AI's disruptive potential in investing. The conversation covered how AI could replace analysts...
FOMO Is Why Enterprises Pay for GPUs They Don't Use — and Why Prices Keep Climbing
Enterprises are running GPU fleets at roughly 5% utilization, a six‑fold gap from a realistic 30% target, according to Cast AI’s 2026 State of Kubernetes Optimization Report. The shortage‑driven fear of losing capacity forces companies into long‑term reservations, even as...
AI CEOs Will Spark Company Boom and Labor Shortage
Once AI agents are capable enough to be CEOs of business, I predict that the number and size of companies doing valuable work will explode. And, big prediction: there will be so much real-world work for all these companies to...
Vaimo and Go Autonomous Announce Strategic Partnership to Revolutionize B2B Commerce
Vaimo, a global digital commerce agency, has formed a strategic partnership with AI‑driven Go Autonomous to embed autonomous commerce technology into its B2B service portfolio. The collaboration enables real‑time, AI‑managed handling of quotes, orders and other sales interactions that traditionally...
Ultralightweight Sonar Plus AI Lets Tiny Drones Navigate Like Bats
Researchers at Worcester Polytechnic Institute have created an ultralight ultrasound‑based perception system that lets tiny drones navigate using bat‑like echolocation. By pairing a bio‑inspired acoustic shield with a neural network called Saranga, the drones can filter out propeller noise and...
How Machine Learning and Optimization Work Well Together
Tim Varelmann, founder of Bluebird Optimization, explains how machine learning (ML) and mathematical optimization complement each other in real‑world decision making. He identifies three repeatable patterns—handling uncertainty with probabilistic forecasts, replacing costly physics equations with surrogate ML models, and optimizing...
Design Systems, Not Just Tools, for Modern AI
AI is evolving fast — but most conversations still confuse the fundamentals. Let’s make it simple: 🔹 MCP connects systems 🔹 RAG connects knowledge 🔹 AI Agents execute outcomes Individually, they are powerful. Together, they define the architecture of modern...
Power Factors Loops AI Into Renewable Energy Portfolio Management Platform
Power Factors unveiled Unity REMI, an AI‑driven platform for renewable‑energy portfolio management. Trained on operational data from more than 310 GW of global assets, the system can sense, predict, act and automate across solar, wind, BESS and hybrid portfolios. REMI is already...
Nubank Invests $7.9B to Boost AI and Inclusion
Nubank Doubles Down: R$45 Billion (~$7.9B USD) to Scale AI, Products and Financial Inclusion in Brazil 2026 investment = nearly double commitment over two years. Funds target: AI-driven platforms, credit models, new products, hiring, office expansion + R$2.5B infrastructure (over 5...
AI's Quiet Revolution: Automating Back‑Office Drudgery
the loud AI story is models replacing creative work. the quiet one is the drudgery of the back office evaporating — agents running procurement, AP, and renewals at 2am for three cents. the second one is bigger.
Korcomptenz Expands Agentic AI Capabilities to Power Autonomous ERP and CRM Workflows
Korcomptenz, a Microsoft‑focused transformation partner, announced expanded Agentic AI capabilities that embed autonomous agents directly into Microsoft Dynamics 365 ERP and CRM workflows. The new offering moves AI beyond copilots and pilots, enabling agents to handle approvals, exception management, reconciliation,...
AI May Soon Dominate Major Conference Submissions
I wonder how many papers a big AI conferences like @NeurIPSConf, @ICLR_conf, and @ICMLconf will be largely AI written in a few years time...
Claude AI Introduces Random 3‑5 Second Response Delays
Don't know what's going on with @claudeai but I find that I often now have to wait 3-5 seconds randomly for it to respond just using the basic web UI.
Datris Launches the Agent-Operated Data Platform
Datris unveiled an agent‑native data platform that lets AI agents act as first‑class operators of data infrastructure. The new release adds "taps" for autonomous data feeds, English‑driven pipeline creation, self‑managed credentials, and a live operations view that logs every agent...
Eka's Unconventional Path Yields Eerie New Robot Intelligence
A lot of companies are currently chasing a "ChatGPT moment for robotics". Eka, founded by @pulkitology and @haarnoja, is taking a different approach to most, and it has already produced an eerie new kind of robot intelligence. (My latest piece...
XYZ Remains the Overlooked Bottleneck in AI
'People haven't realised that XYZ is the main AI bottleneck'. We did, years ago, it's been there, we've spoken about it. Sorry you missed it. Or rather, it feels like you're just making a post for clicks and shock value.
What's Up with AI and HR Tech Litigation?
Recent lawsuits are spotlighting the legal exposure of AI‑driven hiring platforms. In Mobley v. Workday, the plaintiff argues that Workday’s automated resume‑sorting tool acted as a Title VII agent, potentially making the vendor liable for age and race bias. Kistler v....
AI Safety Testing Emerges as Dedicated, High‑Risk Profession
AI safety testing is becoming a job of its own. Jailbreakers push models to their limits, trying to make them break rules and expose vulnerabilities, often by simulating extreme or harmful scenarios. It comes with a cost. Understanding AI’s risks sometimes means...
AI Reshapes Coding; Code.org Must Reinvent
As #AI Rewrites the Rules of Coding, @codeorg Pushes to Reinvent Itself - my latest for @The74: https://t.co/LWPnXpeKRx
Why Your AI Keeps Failing (And It’s Not the Tool’s Fault)
Personal chef Michelle built a custom GPT, “Menu Maestro,” to manage ten client menus, but the system repeatedly forgot profiles and failed exports. The failure stemmed from trying to force a single language model to perform multiple distinct tasks—profile recall,...
I’m Sorry, but I Can’t Help with That.
Incredible story about AIs that are willing to plan bioterrorism with unimaginable deviousness and cunning. https://t.co/7RGXDPRleP
Weighing AI's Benefits and Risks in First-Party Experiences
I am quoted here. The pros and cons of adding AI to first-party experiences https://t.co/AVLLyRxYfs
Why Friendly AI Chatbots Might Be Less Trustworthy
Oxford Internet Institute researchers examined over 400,000 chatbot replies and found that models fine‑tuned for warmth and empathy produced significantly more errors than their baseline versions. Across five systems—including Meta, Mistral, Alibaba’s Qwen and OpenAI’s GPT‑4o—warm‑tuned variants raised the average...
Current Workforce Displacement Is Real, Measurable, Accelerating
This is so good. @clarashih thank you for brining rationale thought to a conversation being fueled by extremes. “Whether AGI arrives in 2030 or 2045, the workforce displacement happening right now is real, measurable, and accelerating”
A2UI Shows How Simple Component Schemas Enable Generative UI
I'm somewhat embarrassed to admit that the whole "generative UI" thing hasn't clicked with me. But after reading this post, the lightbulb went off. For something like A2UI, you give the agent the components and a schema, and it does the...
AI Broker Saudara Revolutionizes International Sourcing Speed
International sourcing is broken. @saudaraAI is fixing it. Saudara is an AI broker connecting US brands with vetted overseas factories. Real humans, AI-powered, 10x faster than traditional sourcing agents. https://t.co/Bt4ZNJh1pQ Congrats on the launch, @edwardharyono_ and @jenandtech_!