Today's AI Pulse
OpenAI unveils ChatGPT Images 2.0 with multilingual text and ‘Thinking’ mode
OpenAI rolled out ChatGPT Images 2.0, a next‑generation image model that adds multilingual typography, real‑time web research, and agentic reasoning. The new ‘Thinking’ mode lets paid users generate up to eight coherent images per prompt, including floor plans, maps and UI mock‑ups. The feature expands the model’s ability to work with uploaded content.
Also developing:
New National AI Facility Under €17 Million Research Ireland Investment
The Walton Institute at South East Technological University secured over €1 million (≈$1.1 million) as part of a €17 million (≈$18.5 million) national Research Ireland programme to build a new AI and high‑performance computing facility. The centre will expand the existing data‑centre capacity by up to 100 times while improving energy efficiency and linking to the European Open Science Cloud. It will serve researchers, SMEs, industry and public‑sector partners across healthcare, climate, agriculture, finance and security. The investment positions SETU as a national AI hub and bolsters regional growth in the South East.
How One Company Is Using Pay Raises to Drive AI Adoption
Omnisend, a marketing automation platform with about 250 staff, is tying annual salary increases to employees' use of artificial intelligence. Workers can earn an extra 2% to 4% raise based on how they integrate AI tools into their daily tasks....
Anthropic’s Project Glasswing Is a Warning: Technical Debt Is Now a National Security Risk
Anthropic unveiled Project Glasswing, an AI model named Claude Mythos that can autonomously discover and chain high‑severity software vulnerabilities. The rapid rollout sparked urgent discussions among the U.S. Treasury, the Federal Reserve, and major banks about the fragility of legacy...
Growing AI Power Slurpage Prompts MPs to Examine Low-Energy Computing
British MPs have launched a short‑term inquiry into low‑energy computing to curb the soaring electricity demand of AI‑driven datacentres. The Science, Innovation and Technology Committee is examining emerging chip designs such as neuromorphic computing and silicon photonics, which promise far...
AI Overhyped Now, Under‑Hyped for Future Gains
Stanley Druckenmiller on AI in late 2024: "If we were all sitting here in 1999 talking about the Internet, I don’t think anybody would have estimated it would be as big as it got in 20 years. And yet, if...
Four-Month-Old Recursive Superintelligence Raises $500m
Recursive Superintelligence, a London‑based AI startup founded four months ago by UCL professor Tim Rocktäschel and former Salesforce chief scientist Richard Socher, announced a $500 million Series A round at a $4 billion valuation. The financing was led by Google Ventures and included...
I Built an AI App with a Tiny Team. We Scaled It to Millions and Sold It to MyFitnessPal in...
Cal AI, an AI‑powered calorie‑tracking app founded by Jake Castillo and three teammates in April 2024, grew to millions in monthly revenue and was acquired by MyFitnessPal in under two years. The startup’s growth hinged on aggressive influencer marketing, signing...
Coinbase Testing AI Agents on Slack and Email
Coinbase has begun testing two internal AI agents that appear in Slack and email, each modeled after former executives Fred Ehrsam and Balaji Srinivasan. The Ehrsam‑based agent provides strategic guidance, while the Srinivasan‑based agent is tasked with sparking innovation. CEO...
Self‑driving Cars Now Infer Intent, Not Just Objects
A Xiaomi car reportedly flagged a nearby Range Rover as likely undercover police. And that is a much bigger signal than it sounds. What interests me is not whether it got this one case perfectly right. It is what this points to. Cars are...
Docplanner Expands Patient Access with Voice AI Agent Powered by Twilio
Docplanner, a leading European‑Latin American health platform, has launched Noa Booking, an AI‑powered voice agent built on Twilio’s ConversationRelay and Programmable Voice. The 24/7 agent lets patients schedule appointments, get FAQs answered and receive instant SMS/WhatsApp confirmations, eliminating traditional call‑centre...
Copilot Exposes Messy Data—Clean up Before Deployment
"Copilot only returns content the user already has access to." That sentence is doing a lot of work right now. Every stale M365 group, every "Anyone with the link" doc — one prompt away now. Copilot didn't create your data mess. It made...
Has AI As A Writing Partner Been Oversold?
Educators once hailed AI as a writing tutor that could clean up ideas without taking over authorship, but Erik Ofgang argues the reality is the opposite. While AI tools like ChatGPT have markedly reduced grammatical errors, they have also made...
Chinese Tech Workers Are Starting to Train Their AI Doubles–And Pushing Back
Chinese tech firms are urging employees to train AI agents that can mimic their coworkers, spurring a viral reaction to the GitHub‑hosted Colleague Skill tool. The spoof project automatically harvests chat histories and files from Lark and DingTalk to generate...
Contracts Are Turning Into Budget Protectors with Spend Intelligence
Financial contract intelligence, an emerging AI category, converts static vendor contracts into structured, queryable financial data. By parsing pricing, terms, and obligations, platforms like SpendBrain enable finance teams to validate charges, spot overpayments, and proactively manage renewals. This transforms spend...
Balancing Health AI Management with Growing Vendor Sprawl
Health AI spending in the United States surged to roughly $1.4 billion in 2025, prompting hospitals to adopt dozens of niche AI tools. A 2026 survey of senior tech leaders found 69% cite vendor management and integration as the biggest barrier,...
3 Questions Every Payer Should Ask About Medical AI
Recent research shows 85% of healthcare leaders expect AI to reshape clinical decision‑making within five years, yet fewer than half of payers have a formal AI strategy. Compliance concerns—particularly around transparency, bias mitigation, and physician oversight—are driving tighter regulatory expectations...
AI in Parcel Shipping: How to Cut Through the Noise and Find Tools that Actually Work
The article warns logistics firms to evaluate AI tools for parcel shipping based on problem relevance, domain specificity, data security, and integration. It stresses that AI should reduce cost per package, prevent lost shipments, and improve carrier management. Vendors must...
AI and Libraries: Why Librarians May Become Arbiters of Reality
A recent Book Industry Study Group webinar revealed that librarians are the publishing industry’s early‑warning system on AI. A BISG survey shows roughly one‑third of library respondents neither use nor plan to use AI, and 34% label themselves ethically opposed....
Powering Partner Success with Innovation and AI: Avalara November 2025 Partner Webinar Highlights
Avalara’s November 2025 partner webinar unveiled an agentic AI strategy that embeds artificial intelligence across its compliance platform. New capabilities include AI‑driven reporting, plain‑language tax research, and integrations that let partners connect their own AI tools via Google’s A2A protocol....
#356 The Forecast for Time Series Forecasts with Rami Krispin, Senior Manager of Data Science at Apple
In this episode, Rami Krispin, Senior Manager of Data Science at Apple, discusses the emergence of time‑series foundation models and how they enable forecasting at massive scale, such as for thousands of SKUs in retail. He contrasts traditional statistical approaches...
3 Ways AI Is Humanizing Patient Care
A 2024 AMIA survey found that over 74% of clinicians say documentation tasks impede patient care, prompting hospitals to turn to AI solutions. Microsoft’s Dragon Copilot is being deployed at Cooper University Health Care, Mercy’s Fort Smith hospital, and the University...
Making AI Actually Work in the Enterprise and More RSAC Conference 2026 Interviews - Aamir Lakhani, Camellia Chan, Ely Abramovitch,...
The episode explores why many enterprise AI initiatives stumble, emphasizing the need for security‑by‑design and clear governance. Fortinet’s 2026 Global Threat Landscape Report highlights a sharp rise in AI‑enabled cybercrime, prompting vendors like X‑PHY to deliver hardware‑enforced safeguards for AI...
Smarter Stores Start Behind the Scenes
Retailers are re‑thinking the hype around customer‑facing AI and realizing that the true value of artificial intelligence lies in behind‑the‑scenes operations. Large language models work best when tightly tailored to interpret and act on internal data, turning fragmented information into...
Shadow Data Is a Hidden Risk that AI Can Reveal
Shadow data—forgotten copies of sensitive information scattered across endpoints, cloud storage, SaaS tools and generative‑AI prompts—remains invisible to most security teams. Research from Palo Alto Networks shows over 80% of sensitive data lives outside traditional controls, a problem amplified by...
With AI Entering Bidding Workflows, the Estimator’s Role Isn’t Shrinking; It’s Expanding
AI is reshaping construction estimating by automating drawing analysis and quantity takeoffs, turning a traditionally manual process into a rapid, data‑driven workflow. With U.S. construction spending approaching $2.2 trillion, the sector faces a shortage of over 300,000 skilled workers, stretching estimator...
Connected Data, Connected Decisions: The Real AI Advantage for Inventory Leaders
Dealers are drowning in data but lack a unified infrastructure, limiting AI’s effectiveness. Cox Automotive executives highlighted at NADA that connecting DMS, CRM, appraisal and market data transforms inventory decisions, enabling precise acquisition, channel prioritization and sharper appraisals. A case...
QNX Integrates with Nvidia IGX Thor to Accelerate Safety-Critical AI Deployment
At Hannover Messe, QNX, a BlackBerry division, announced an expanded partnership with Nvidia to integrate QNX OS for Safety 8.0 into Nvidia's IGX Thor platform and its Halos Safety Stack. The unified solution combines QNX's deterministic real‑time operating system with Nvidia’s functional‑safety‑rated...
Functional AI For 3D Design Automation — From Path Finding To Generative Modeling For Building Construction
In this talk, Richard, a professor at Simon Fraser University and VP of AI at the startup Augmenta, explores the frontier of functional AI for 3D design automation in construction. He reviews his academic work on geometry decomposition, Fermat-spiral path‑filling,...
NVIDIA and Partners Showcase the Future of AI-Driven Manufacturing at Hannover Messe 2026
At Hannover Messe 2026, NVIDIA and partners demonstrated a full suite of AI‑driven manufacturing solutions, from accelerated computing infrastructure to real‑time digital twins and autonomous robots. The Industrial AI Cloud, built by Deutsche Telekom on NVIDIA hardware, provides a sovereign,...
UiPath’s Next Chapter in Retail: From Peak’s AI Heritage to Agentic Merchandising at Scale
UiPath’s acquisition of Peak in March 2025 brings a decade of retail AI expertise into its automation platform, enabling what the company calls "agentic AI" – AI that not only predicts but also executes merchandising decisions. The combined solution links demand...
AI Model Claude Opus Turns Bugs Into Exploits for Just $2,283
Anthropic's Claude Opus 4.6 model produced a functional Chrome V8 exploit after processing 2.3 billion tokens across 1,765 API calls, costing roughly $2,283. The experiment demonstrated that publicly accessible AI can turn software patches into weaponized code with modest investment. Researchers...
AI Scribes Add Modest Revenue, Uncertain Burnout Impact
Studies keep coming out showing how much time AI scribes help save for physicians. But how about return on income? As AI scribes usually lead to modest reductions in EHR time and documentation time, they also come with a small but...
Denis Beau: The Challenges Posed by AI From the Perspective of the Central Bank
Denis Beau, senior official at the Banque de France, warned that artificial intelligence is reshaping the central bank’s core mandates of monetary and financial stability. He emphasized the need to monitor AI’s diffusion across the economy and its ambiguous effects...
Here's Why Most AI Initiatives Crash at Pilot Stage
Enterprises are rushing into generative AI, yet an MIT study shows roughly 95% of pilots never reach production and deliver no measurable ROI. Experts like SS&C Blue Prism CTO Lou Bachenheimer argue the failures stem from mismatched use cases, lack...
A Formal Model of How Artificial Intelligence Erodes Human Agency
Researchers Alvin Moon and Benjamin Boudreaux present a formal model that quantifies how artificial intelligence erodes collective human agency. Using social‑choice theory, they introduce three metrics—distribution of decisive coalitions, minimal coalition size, and composition of minimal coalitions—to track shifts in...
A Strange Quirk of the Legal Profession Means Lawyers May Soon Have to Adopt AI—Or Face Malpractice
Lawyers are facing mounting pressure to integrate artificial intelligence after recent disciplinary actions highlighted the risks of AI hallucinations. A Massachusetts attorney was sanctioned for citing fabricated cases generated by ChatGPT, and a California lawyer was fined $10,000 for similar...
How to Reverse-Engineer the Perfect ChatGPT Prompt, According to an MIT Professor
MIT’s Andrew Lo, director of the Laboratory for Financial Engineering, outlined a systematic approach to crafting high‑quality prompts for AI chatbots such as ChatGPT, Gemini, and Claude. He emphasized that vague queries produce generic answers, while detailed, context‑rich prompts unlock...
AI Could Counteract Aging Populations' Economic Strain
AI is entering an unexpected conversation around demographics. Falling fertility and aging populations are already straining economic systems, and AI could either offset the impact through productivity gains or deepen inequalities if benefits are uneven. The question is broader than technology. It...
Peter Stetson, Tiger Connect
Tiger Connect, now branding itself as an AI platform for clinical communication, aims to orchestrate messages between doctors, nurses, EMTs and connected devices. CMIO Peter Stetson, formerly of Memorial Sloan Kettering, says the core challenge is routing alerts to the right person...
Medicine Is a Flowchart; AI Outperforms Doctors
The dirty secret no doctor wants to admit: Medicine is just a flowchart. Symptom → Rule → Drug → Done. AI doesn't get tired. Doesn't miss patterns. Doesn't have a bad day. So seriously — why do you think your MD...
Goldman Sachs Views AI as Evolution, Not Apocalypse
The 10-Year Pivot: Why Goldman Sachs Thinks AI is an Evolution, Not an Apocalypse by @Timothy_Hughes https://t.co/gnz4oSh31V @DLAIgnite #Sales #SalesTips #SalesLeader #Marketing #MarketingStrategy #Leadership #ArtificialIntelligence #Innovation #Technology #Tech #TechNews https://t.co/X1m0xwvDq7
Understanding the Data: AI & Workforce Readiness
DataCamp’s 2026 State of Data & Literacy Report reveals a widening gap between AI expectations and workforce readiness. While 90% of senior leaders rank data literacy above traditional skills, 60% admit internal AI and data skill gaps and fewer than...
AI Delivers Unprecedented Accuracy in Demand Forecasting
Supply chain managers long for accurate demand forecasts. AI is making quantum leaps in this area, processing vast internal & external data for predictions beyond human capability. #AISupplyChain #DemandForecasting https://t.co/Fx6oi6tyJx
Avoid Geographic Silos: Foster Free AI Idea Exchange
The AI Science Separation “It’ll be a massive loss if researchers get into geographical siloes and don’t freely cross pollinate ideas,” Rao says. No kidding... https://t.co/VvWR05BnZR
Hotspring Upgrades AI-Powered VFX Interface
Hotspring unveiled version 2.0 of its AI‑powered VFX platform, consolidating production tools into a single interface and adding project‑wide thumbnails with full metadata. The upgrade expands in‑house AI capabilities for roto, paint and tracking, enabling a hybrid workflow where AI generates...
AI-Driven Drug Discovery: Current Landscape and Future Promise
Cool review and perspective on the current state and the future of AI-powered drug discovery https://t.co/AgwqOlt4RU
NIH-Supported Study Develops AI Algorithm Trained on EHR Data to Predict Rare Disease
A NIH‑supported study unveiled the WEakly Supervised Transformer (WEST), an AI model that can predict rare diseases using noisy, incomplete electronic health record data. Tested on pulmonary hypertension and severe asthma, WEST outperformed all baseline models in identifying clinician‑confirmed cases....
Makko Launches AI-Powered 2D Game Studio with over 40,000 Game Assets Created in Beta
Makko AI has officially launched its AI‑powered 2D game studio at makko.ai after a beta focused on quality and creator feedback. The platform lets users generate cohesive 2D art assets and playable prototypes through conversational prompts, eliminating traditional drawing, animation,...
Tennessee Becomes Another State To Enact New Law That Restricts AI Acting As A Mental Health Advisor
Tennessee Governor Bill Lee signed Senate Bill 1580 on April 6, 2026, adding Section 33‑1‑205 to the state code. The one‑page law, effective July 1, 2026, prohibits developers and deployers from advertising AI as a qualified mental‑health professional. Violations are treated as breaches of the...
BingX Reports Strong Q1 2026 Growth as AI Users Surpass 5 Million and TradFi Hits 50% of Volume
BingX launched its TradFi Market in Q1 2026, offering over 100 traditional assets and capturing 50% of the platform’s trading volume. The exchange’s AI suite expanded with AI Claw and the AI Skills Hub, driving AI user numbers past 5 million...