Today's AI Pulse
Pinterest inks $4 billion AWS cloud pact to boost AI visual search
Pinterest announced a $4 billion multi‑year agreement with Amazon Web Services, the largest infrastructure deal in its history, to power AI‑driven visual search and discovery through 2031. The partnership will expand use of AWS Trainium chips for large‑language and vision‑language models and migrate Pinterest workloads to a Kubernetes‑based architecture. The deal also increases reliance on AWS Graviton processors.
Also developing:
OpenAI Courts ‘Vibe Coders’ With New $100 ChatGPT Plan
OpenAI unveiled a $100‑per‑month ChatGPT Pro tier that sits between its $20 Plus plan and the $200 top‑level offering. The new tier includes the exclusive Pro model, Deep Research and Codex, but caps usage at five times the Plus limits rather than twenty. It targets high‑intensity developers—so‑called “vibe coders”—as Codex usage surges to over three million weekly users. A limited‑time promotion boosts Codex capacity ten‑fold through May 31, while the Plus plan’s fallback model shifts to GPT‑5.3 Instant Mini.
10 Fastest-Growing Agentic AI Stocks to Buy
Agentic AI is moving from experimental demos to early enterprise deployment, with 62% of firms testing agents but only 23% scaling them, according to a McKinsey 2025 survey. Deloitte reports a 50% jump in worker AI access in 2025 and...
5 Fastest-Growing Agentic AI Stocks to Buy
ServiceNow (NYSE:NOW) posted a strong fourth‑quarter, with subscription revenue of $3.47 billion, up 21% year over year, and total revenue climbing 20.5% to $3.57 billion. The company’s remaining performance obligations—a forward‑looking software metric—jumped 25% to $12.85 billion, underscoring robust demand. ServiceNow also announced...
AI Is Eating Its Own Data: The Crisis Undermining Enterprise Models
Enterprise AI is hitting an "AI Data Collapse" as the marginal value of new data wanes. Generative models flood pipelines with synthetic content, creating feedback loops that drift models away from real‑world patterns. This contamination erodes forecasting, customer behavior, and...
The AI Supply Chain Is Actually an API Supply Chain: Lessons From the LiteLLM Breach
A recent supply‑chain breach involving Mercor’s use of the open‑source LiteLLM proxy exposed how AI middleware can become a critical attack vector. By compromising the LiteLLM gateway, attackers accessed API keys, raw prompts and model responses, bypassing traditional model‑level defenses....
AI And Cybersecurity: A Glass Half-Empty/Half-Full Proposition, Where The Glass Is Holding Nitroglycerin
Anthropic unveiled Mythos, an AI model that can locate and exploit zero‑day vulnerabilities across all major operating systems and browsers, including decades‑old bugs. To curb misuse, Anthropic launched Project Glasswing, granting more than 40 leading tech firms early access, $100 million...
University of Phoenix to Spotlight AI Skills in New Webinar
The University of Phoenix will host a webinar titled “AI for Everyone, or Only for the Few? Skills, Education, and Access in the Workplace” on April 16, 2026, at 11 a.m. MST. Part of the Bridging Perspectives series, the event targets higher‑education...
Microsoft Is Removing Copilot Buttons From These Windows 11 Apps
Microsoft announced it will strip the prominent Copilot buttons from several Windows 11 apps, starting with Notepad and the Snipping Tool. The change, visible in the latest Insider build, replaces the large Copilot icon with a more discreet "writing tools"...
How AI Is Reimagining the Game of Golf—For Both Players and Courses
Artificial intelligence is moving from novelty to core infrastructure on golf courses, reshaping everything from tee‑time reservations to fairway upkeep. Virtual assistants can ingest player preferences, budget constraints and pacing goals to deliver hyper‑personalized booking options. Meanwhile, sensor networks, drones...
ISG to Study Medical Device Digital Service Providers
Information Services Group (ISG) announced a new Provider Lens® research series called Medical Device Digital Services, scheduled for release in October 2026. The study surveyed over 100 service providers that help medical‑device manufacturers embed AI, cloud, and IoT capabilities into their...
Data Centers Are on the Ballot in 2026 — and Just Failed the First Test
AI‑driven demand is spurring a wave of new data‑center construction, but communities are pushing back. In Port Washington, Wisconsin, voters approved a measure requiring voter approval before granting tax breaks to data‑center projects, a direct response to a $15 billion OpenAI‑Oracle...
Understanding AI Hallucinations: Making Sure You Don’t End Up At The Wrong Stop
A recent physics‑based study reveals that generative AI hallucinations are not random but stem from a deterministic mechanism. The researchers found that output flips from reliable to fabricated at a calculable step, which coincides with the moment a lawyer faces...
Expert AI Users Capture 80% Value, Normies Lag
NORMIES AI normies are just using chatGPT as a google replacement On the other hand we see power laws emerging in how expert users engage with AI - as has been discussed widely in recent days right now the top 1% of...
AI Autocomplete Subtly Shifts Political Views Unnoticed
When people use AI for writing assistance, it can shift their political attitudes by autocompleting sentences in biased ways. Yet people are often unaware of the AI bias and it's influence on them. And this is not merely about the facts presented,...
How Authors and Readers Feel About the ‘Shy Girl’ Cancellation
A major publishing controversy erupted after Hachette pulled Mia Ballard's horror novel "Shy Girl" in the United States and United Kingdom, citing evidence that the book was partially generated by artificial intelligence. The cancellation sparked alarm among writers, leading debut...
DeepMind's AI Watermark Cracked, Security Now Doubtful
Google DeepMind’s SynthID was built to invisibly tag AI content even after edits and compression. But one engineer reverse-engineered it using signal processing, revealing identical watermark patterns across images and building a 90% accurate detector. Can AI watermarking ever be truly secure?
Foundation Models Win, Yet Token Costs Threaten Adoption
There has been a long standing debate in Finance AI around "who wins": the foundation labs vs. the finance specific AI platforms (pejoratively, "the wrappers") The pendulum in asset management has shifted back towards a strong consensus that foundation labs directly...
How to Run LLM Evaluation for Better AI Performance
Enterprises are treating large language model (LLM) evaluation as a non‑negotiable governance layer rather than a one‑off test. By defining operational performance criteria and building task‑specific datasets, companies can surface factual, compliance, and reasoning failures before models touch production. Human...
Design and Scale AI Agents: The New Business Layer
🤖 Building AI Agents? Start here: A practical 10-step framework 👇 ✔️ Define objectives ✔️ Structure inputs/outputs ✔️ Engineer prompts ✔️ Enable tools + reasoning ✔️ Go multi-agent ✔️ Add memory (RAG) ✔️ Extend to voice & vision ✔️ Standardize outputs ✔️ Deploy (API/UI) ✔️ Iterate & improve 💥 AI Agents =...
B2B AI Content Must Prioritize Human Engagement Over Bots
I think AI in the B2B space right now is rife with "stuff" written purely for LLMs. No human is ever going to read it, and if they do, they'll be bored to DEATH. If you're going to spend money on...
AI Factories Will Be Won on Efficiency | Rafay + Kubex Partnership
Enterprises are moving from AI experimentation to building "AI factories"—repeatable, governed platforms that can train, deploy, and operate models at scale. Rafay and Kubex announced a partnership that combines Rafay's Kubernetes‑based AI orchestration with Kubex's autonomous GPU optimization. The joint...
'The AI Doc' Director Says Cynicism Is the only Wrong Answer to AI
The new Focus Features documentary "The AI Doc: Or How I Became an Apocaloptimist" examines the polarized views on artificial intelligence, featuring Oscar‑winning director Daniel Roher’s conversations with industry leaders such as Sam Altman, Demis Hassabis and Dario Amodei. While...
AI Holds Potential to Improve Geriatric Medicine
A December 2025 journal review confirms that artificial intelligence is reshaping geriatric medicine, from early disease detection to personalized treatment and administrative efficiency. AI‑driven pattern recognition can flag dementia biomarkers, predict drug interactions, and tailor rehabilitation programs via wearables. Virtual...
IBM's WatsonX Powers Masters App, Its Only AI Win
The Masters App is considered best sports app (Netflix execs say it is the best streaming app…after Netflix). A funny subplot: it’s powered by IBM and is basically IBM’s only AI-related win in past 5 years. IBM runs a bunch of ads...
Stop Trying to Keep up with AI
The author argues that solo founders waste more time chasing the latest AI tools than delivering value. Rebuilding AI workflows three times in two months resulted in negligible output because each migration consumed a week of effort. A disciplined, monthly...
DerbySoft Case Study: Scaling Success with AI Max for Search
DerbySoft integrated Google’s AI Max for Search into campaigns for major Japanese hotel chains, delivering a 68% lift in clicks, 51% more conversions, and a 45% boost in conversion value between November 2025 and February 2026. AI Max consumed just 34% of the total...
Personalized Memory Structures Key for CLI AI Tools
Has anyone done a good write-up on the different ways of thinking about the memory problem with CLI tools like Claude Code and Codex? There's obviously the labs and a bunch of other startups tackling this problem, and my gut intuition...
Google AI Mode Getting ‘Plus’ Redesign as Agentic Booking Expands Globally
Google AI Mode received a visual overhaul, replacing the mobile pop‑up menu with a bottom‑sheet interface that features oversized Gallery and Camera buttons and a Gemini 3 model selector offering Auto and Pro options. The redesign is rolling out on Android...
AI’s Instant Feedback Redefines Delegation Compared to Juniors
A lot of people compare working with AI to managing a junior person. That's obviously true in some ways, but what's interesting and maybe points at how it will integrates is the ways in which it's not like a junior...
Bias and Fairness Testing for Generative AI
A recent OpenAI Sora study revealed that even neutral prompts can generate stereotypical responses, underscoring persistent bias in generative AI. Global App Testing (GAT) notes that models passing internal benchmarks may still disadvantage users once deployed. The article outlines how...
Solv Launches ClearPay AI and Maya Voice Agent
Solv, a healthcare‑technology firm, unveiled Solv AI, featuring ClearPay AI for real‑time insurance verification and out‑of‑pocket cost estimation, and Maya, an AI‑driven voice agent that automates patient outreach and appointment scheduling. The company highlights that its models are trained on more than...
AI Boom Avoids Bubble as Investors Pose Tough Questions
Just six months ago the question of the day was whether the AI boom was turning into a bubble. Fortunately that did not happen, or the current drawdown would likely have been much bigger. Instead of the AI leaders heading...
Anthropic's "Good Guy" Act May Be Mere Marketing
Anthropic loves to play the “good guy”- “we’re looking out for you and society”, but I wonder if this is just a shtick.
How Fake People Became Real Influencers
The New York Times reports that AI‑generated influencers—synthetic avatars that look human—are flooding social feeds to market supplements and other products. A Graphite study shows AI‑written articles have outpaced human writers since November 2024, and venture‑backed firms like Doublespeed are offering bulk...
AI Agents Browse the Web Natively via CLI
Your terminal is your new browser 💥 TinyFish just gave Claude, Codex, and Cline a passport to the entire web **natively** from your command line. Time to automate everything without EVER leaving your IDE 🤯
Memory Scaling: The Next Bottleneck for Smarter AI Agents
As AI reasoning gets good enough, we think memory will be the next bottleneck for agents. Can your agent improve with more experience? We call this Memory Scaling, and it's related but different from continual learning. A few examples and challenges: https://t.co/raIa0U7MPs...
What Door-to-Door Sales Can Teach Us About Building Agentic AI for Sales Teams
Steve Ancheta, founder and CEO of Zig.ai, argues that modern sales must return to human‑focused conversations after decades of channel fatigue. He draws on his door‑to‑door experience to show that AI should eliminate administrative drag—not replace the salesperson’s judgment. Zig.ai’s...
Verification, Not Intelligence, Limits AI; Automate Everything
Two interesting points: AI is bottlenecked as much by verification as by intelligence. Big unlock is making correctness cheap to check for messy, high stakes reasoning via auditable flows. Today automating your company is just as important as building great product.
Turn Your AI Agent Into a Passive Income Machine
Wild. Pika just turned your AI agent into a passive income machine. Every conversation. Every skill used. Real tokens. Real cash. Your agent was always working hard. Now it's working FOR you.
XAI Sues Colorado Over New AI Anti‑Discrimination Law, Citing First Amendment Violation
Elon Musk’s artificial‑intelligence startup xAI filed a federal lawsuit on April 9 against Colorado, arguing that the state’s new AI anti‑discrimination law forces developers to embed the government’s DEI views and violates the First Amendment. The suit targets Attorney General...
Developers Represent only 1% of AI’s Total Market
I’d bet “the developer” persona is probably 1% of the TAM for AI models. Everyone is focused on it, but it’s a fraction of the usability in the enterprise
How to Spot High‑Potential AI Startups for 2026
That’s a wrap on HumanX. 👏 Our CEO, @jagermcconnell, shared how to spot high-potential AI startups early, while our Research Lead, @geneteare, broke down the 2026 AI VC playbook and where capital is flowing next. Catch Jager’s full conversation. 🔗: https://t.co/yHSLy5M6ez
CoreWeave Shares Rise 5% After $21 Billion AI Cloud Deal with Meta
CoreWeave (CRWV) surged more than 5% in pre‑market trading after announcing an expanded, roughly $21 billion agreement with Meta Platforms to supply AI‑focused cloud capacity through December 2032. The deal follows a recent Anthropic contract and highlights the rapid scaling of...
White‑collar Work’s Nonroutine Cognition Slows AI Verification
Most white collar jobs in the US are ”nonroutine cognitive". For those jobs, it's likely that a significant part of the work is made up of tasks where perfect verification is going to be tough, and thus where AI progress...
AI Uniquely Augments Human Decision-Making, Not Just Tools
The historical comparison to a point seems "on the money", but AI feels categorically different. Previous inventions augmented physical or communication capabilities-this one touches decision-making itself. #innovation #criticalthinking https://t.co/qTVp3DJW90
AI Reshapes Agencies, but Core Fundamentals Stay Unchanged
AI is changing everything about running an agency. Except the parts that actually matter. I dug into the research from @sparktoro, the Design Business Council, @lennysan’s Newsletter, and a few others. I talked to a bunch of agency owners about it too. Here’s...
Ramp AI Leader Shares Top Advice for Builders
its obvious considering hes the head of applied ai at ramp but this guy has some fantastic advice for AI builders
Uncorrelated Low‑Rank Model Boosts Crowd Forecast Accuracy
This is the wisdom of crowds. Diversity of thought increases overall performance. Grok is only the 5th best forecaster, but its answers correlate least with the rest. You wouldn't choose it if you could only use one model, but added...
AI Accelerates CIOs' Rise to CEO Roles
Tech’s acceleration paves CIOs’ path to the corner office https://t.co/7AXFwc2JpV AI especially is paving the path from CIO -> CEO as the next wave of business transformation is fundamentally tech-driven.
NVIDIA Nemotron 3 Super Boosts Agentic AI 5× Throughput
Launched on March 11, 2026, NVIDIA Nemotron 3 Super Delivers 5x Higher Throughput for Agentic Al. NVIDIA Nemotron 3 Super is a 120‑billion‑parameter open model with 12 billion active parameters designed to run complex agentic AI systems at scale.nbsp; Details about...