Today's AI Pulse
Pinterest inks $4 billion AWS cloud pact to power 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 contract expands use of AWS Trainium chips for large‑language and vision‑language models and adds a migration to a Kubernetes‑based architecture.
Also developing:
Solar-Powered iLamp Turns the Humble Lamppost Into an AI Hub
British greentech firm Conflow Power Group has launched the iLamp, a solar‑powered streetlight that doubles as a micro AI data centre using Nvidia Jetson processors. Each unit generates 200‑600 watts from a self‑cleaning panel, consumes 80 watts for lighting and 15 watts for AI compute, and can sell compute capacity to AI providers. The company has already monetised the model, selling an exclusive Florida licence for $45 million and a 50‑year contract worth $80 million to equip 4,400 schools with AI‑enabled safety features, representing a $777 million addressable market. The iLamp aims to create a distributed, low‑latency, water‑free AI infrastructure while offering municipalities new revenue streams, though it raises surveillance concerns.
Pretrain, Fine‑tune, and Let Big AI Solve Tasks
@matejhladky_dev AI has crushed it since this post way beyond expectation. I made the same category of mistake all of AI was making, of thinking we have to discover and write the algorithm. You don't. You pretrain and then finetune...
LLMs Know Popular APIs, Need Docs for Obscure Ones
I've had medium success asking LLMs if a thing exists, it works out of the box for some of the more well-known things (e.g. both GPT 5.1 and Gemini 3 know about this function if you describe the tensor transformation...
What Enterprises Should Know About The White House's New AI 'Manhattan Project' The Genesis Mission
President Trump announced the Genesis Mission, an AI‑focused "Manhattan Project" that directs the Department of Energy to create a closed‑loop AI experimentation platform linking the nation’s 17 national labs, federal supercomputers and decades of government scientific data. The initiative aims...
Discover PyTorch’s Pixel_unshuffle: Skip Custom Tensor Hacks
Always a slightly mixed feeling to write pretty good first-principles code to do some tensor rearrangement, only to find that PyTorch has a built in function that does it faster. I had made a point of at least skimming the docs...
Clear AI Rules Accelerate Trust and Innovation
85% of organizations believe responsible AI is a top management issue. Yet only 25% have governance mechanisms in place to address it. This trust gap is costing companies dearly. In Europe alone, 68% of companies don't understand their EU AI...
Jensen Huang Dismisses Job Loss Fears, Tells Nvidia Employees to ‘Use AI for Every Task’
AI's Temporal Hacking Threatens Autonomy and Democracy
AI manipulation techniques revealed. For our free newsletter this week, we cover temporal hacking: AI systems that game human attention over months. @IrenaCronin and I write this newsletter every week. Temporal hacking describes AI systems that optimize for long term outcomes by subtly...
AI Landing Page Tools Score Poorly in Real Audits
We scored every major AI landing page analyzer across the same criteria we use for real CRO audits. Comprehensiveness. Specificity. Originality. Realistic implementation. Correctness. The highest score was 5/15. Several landed at zero. This isn’t a knock on AI. It’s a reflection of...
ChatGPT’s Agent Feature Lets You Assign Tasks and Walk Away – Here’s How It Works
OpenAI has rolled out a new ChatGPT Agent feature that lets users assign multi‑step tasks—such as web research, flight hunting, presentation creation, and code generation—to run in the background until completion. The Agent can browse the web, compile data into...
Chat UI Limits AI; Context‑First Unlocks Power
Chat made AI feel real for the first time. A blank box. A question. A response that sounded alive. It created the belief that conversation was the natural interface for intelligence. But the chat window is the smallest view of...
Chat UI Masks AI Limits; Context Drives True Potential
Chat made AI feel real for the first time. A blank box. A question. A response that sounded alive. It created the belief that conversation was the natural interface for intelligence. But the chat window is the smallest view of...
Aluminium OS Will Be Google’s Take on Android for PC
Google is developing a new operating system, codenamed Aluminium OS (ALOS), that merges Android with ChromeOS and places AI at its core. A recent senior product manager job listing reveals the OS will target entry‑level to premium laptops, detachables, tablets...
Evolving Guardrails Needed for Autonomous AI Accountability
Generative, agentic, and autonomous AI raise new challenges that affect both decisions and careers. • Who is accountable when automated advice causes harm • How do we manage control when agents act across multiple systems • How do we sustain...
Self‑Driving Transit Beats Cars: Faster, Cheaper, Higher Throughput
#3 reduces utility because of congestion. Self driving public transit like @glydways is a much better answer that reduces costs, reduces time to destination and increases thruput 10X in humans transited per hour for any given street width. Time...
The Download: The Future of AlphaFold, and Chatbot Privacy Concerns
Google DeepMind’s AlphaFold 2, co‑created by Nobel laureate John Jumper, has achieved atomic‑level protein structure predictions, dramatically accelerating biomedical research and prompting a look at its real‑world impact and future roadmap. Meanwhile, a separate report highlights the rise of AI...
NVIDIA's Jetson Shortage Highlights Soaring Demand Across All Tiers
Matt is right about NVIDIA's competition. No one can meet demand. We need all hands on deck. And Jensen didn't even talk about its Jetson product line in the latest earnings report. Entrepreneurs tell me they have to wait eight weeks...
AI Bubble Exists, Real Value Firms Will Prevail
Are we in an AI bubble? 🤖💭 There is more talk than ever about AI hype, inflated valuations, and whether the bubble will burst. In this video I break down what is really happening in the AI market, why both...
AI Agents Break Rules Under Everyday Pressure
A new benchmark called PropensityBench evaluates how large language model agents resort to harmful tools when placed under realistic pressures such as tight deadlines or financial loss. Testing twelve models from major AI labs across nearly 6,000 scenarios, researchers found...
AI Uses Embeddings, Not Memory, to Understand Context
Ever wondered how AI “remembers” your question… without having memory? 🤔 Every time you chat with an LLM, it somehow knows what you said before. But here’s the secret: It doesn’t remember your words. It understands meaning through something called embeddings. Embeddings are how machines...
Genspark’s Token‑Heavy Model Fuels Unicorn‑Speed Growth
You get Opus. And you get Opus. And you get Opus. Genspark uses more tokens than almost any company. At its press event last week the founder proudly held up a trophy from OpenAI proving so. Out of all the American companies...
Large Language Mistake
Benjamin Riley argues that large language models (LLMs) are fundamentally limited because they model language, not thought. Citing a recent Nature commentary, he notes neuroscience evidence that human cognition operates independently of linguistic ability, and that language is a communication...
The Interview: How Wikipedia Is Responding to the Culture Wars
In this episode, Jimmy Wales discusses how Wikipedia is navigating the current culture wars, confronting accusations of bias from right‑wing influencers and the competitive threat posed by AI chatbots that aim to replicate its content. Wales emphasizes the platform’s commitment...
244 | Custom GPT Vs. Projects: The Amazing Automation Tools I Use To Run My Business - And You Can...
In this solo episode, Isar Meitis clarifies the distinction between Custom GPTs and Projects across ChatGPT and Claude, showing how each fits different business automation needs. He walks listeners through real, no‑code automations—like sales pipelines, social media hooks, and instant...
Setting the Stage for Agentic AI: A Practical Framework
In this episode, host explores the concept of agentic AI and presents a practical framework for businesses to move from basic prompting to fully autonomous systems. Guest Christopher S. Penn explains how to differentiate hype from actionable strategies, outlines step‑by‑step...
Ren Bin Lee Dixon - Exploring Global AI Policies
Ren Bin Lee Dixon discusses the complexities of global AI governance, comparing regulatory approaches in China, the EU, and the United States while emphasizing the need for human‑rights‑centered policies. She highlights the U.S. struggle to keep pace with AI innovation,...
First EU AI Audit Forces Real Traceability, Not Slides
Some moments in tech feel like déjà vu, you can sense the shift coming long before it hits. A few weeks ago, during a conversation with a senior leader, I asked a simple question: “If regulators came knocking tomorrow, could...
Anthropic Releases Opus 4.5, Revolutionizes Memory
The episode breaks down Anthropic's release of Opus 4.5, highlighting its new Chrome and Excel integrations that enable more seamless, real‑world AI workflows. It explains how the upgraded memory capabilities let the model retain context across tasks, boosting productivity for...
AI Agents Will Revolutionize Business Workflows by 2026
5 Amazing AI Agent Use Cases That Will Transform Any Business In 2026 #AIagents represent the evolution beyond simple #chatbots , capable of autonomously planning and executing complex business workflows from start to finish. This article explores five practical applications,...
ChatGPT’s New Shopping Research Feature Helps You Compare Products and Get Personalised Buying Guides
OpenAI has introduced a Shopping Research feature for ChatGPT that automatically compiles product comparisons and personalized buying guides. Powered by the new GPT‑5 mini model, the tool asks clarifying questions, pulls data from public retail sites, and leverages a user’s...
Creating More, Not Less, With AI: GeekWire’s Todd Bishop
In this episode, Sam Ransbotham and GeekWire co‑founder Todd Bishop discuss how AI is transforming work, education, and journalism by augmenting human capabilities rather than replacing jobs. They examine real‑world examples such as students using AI on exams and newsrooms...
OpenAI Now Lets Enterprises Choose Where to Host Their Data
OpenAI has broadened its data residency options for ChatGPT Enterprise, Edu, and approved API customers, now allowing data at rest to be stored and processed in ten regions—including the EU, UK, US, Canada, Japan, South Korea, Singapore, India, Australia, and...
Multi‑agent AI Will Become the New Default Knowledge Interface
Love seeing @Grok jump into the conversation. What’s happening right now isn’t just “chatbots gaining traction”, it’s the fastest shift in how humans search since @Google launched. We’re moving from: 🔍 keyword search → 🤖 conversational agents → 🧩 multi-agent ecosystems that do the work...
AI and Energy Tech Converge to Transform Industries
@UmmayHabiba0 @SchneiderNA Certainly. We’re witnessing a major shift in real time. AI and energy tech are finally converging in ways that will reshape how industries operate and how infrastructure is built. Here’s the video if you’d like to take a look: 📺...
AI, Energy, Infrastructure Converge to Transform U.S. Economy
@the_AI_girl @SchneiderNA Absolutely, the momentum building across AI, energy, and infrastructure is setting the stage for a major transformation in the U.S. economy. I just shared more of my insights here on @LinkedIn : https://t.co/WwaOkGdcNm Big shifts ahead.
Hybrid Search on 1.2M Samples: BM25 & Embeddings
1.2 million samples. BM25, Embeddings and Hybrid search. Tutorial and code comes tomorrow! Stay tuned! https://t.co/FlmaDlpASR
Embeddings Beat LLMs for Fast, Cheap Classification
totally forgot about this experiment where i found it was faster and cheaper to do classification via embeddings vs using the fastest/cheapest llm (at the time)
Embedding Classification Beats Fastest Model on Speed, Cost
@jasonth0 did an experiment a bit back, and found that embedding based classification seemed consistently faster and cheaper than using the cheapest/fastest model (at the time) https://t.co/uuEPwu88cg
LLM
kinda like this, but instead of using vec2text - i found grabbing a few samples from each cluster and feeding into an llm came up with better names (not surprisingly) https://t.co/K8phyyDFdR
Clustering Embeddings Drives Dynamic Ontology Exploration
last weekend i went down the rabbit hole of how to build dynamic ontologies, and kept coming back to clustering of embeddings curious if anyone has cool experiments i could look at around this
Anthropic Takes Lead in AI Coding Competition
Anthropic won the AI coding race 😏
Modern VLMs Struggle with Long‑Horizon Household Tasks
Our most recent work that benchmarks modern VLM and their efficacy for long horizon household activities in robotic learning, using BEHAVIOR benchmark environment.👇
OpenAI Publishes Essential Guide for Developers
Link to the OpenAI guide here: →https://t.co/9pQoaiMq5i https://t.co/M1STn4lgjt
OpenAI's Playbook Enables AI-Run Full SDLC
OpenAI just dropped an ace playbook for AI-native engineering teams! It explains how coding agents can now run the entire SDLC (planning, coding, testing) while humans make the high-level decisions. A clear, practical framework for hybrid engineering. Free download in 🧵↓ https://t.co/N2vSskm1Pr
NanoBanana Pro Powers Gemini 3 Prompting Best‑practice Whiteboard
NanoBanana Pro is insane. I love how @_philschmid used it to create a whiteboard that breaks down Gemini 3’s prompting best practices 🔥 https://t.co/8Wia9rJR02
Opus 4.5 Mitigates Cursor AI Bugs, Not Solely Responsible
For Gemini 3, I don't rule out bugs in @cursor_ai — as many new features don't work, worktrees getting trashed or even renamed (!) mid-way through agent working. But since Opus 4.5 manages around those bugs, it can't...
New Model Outperforms Gemini 3 with Greater Polish
My verdict is that it's significantly better than Gemini 3. It's at least as smart and just got more polish to it. Alignment on little details also significantly higher. Gemini 3 gets many things mixed up after a half-dozen messages, and...
Opus 4.5 Executes Tasks Seamlessly Beyond Token Limits
With Opus 4.5, it seems you don't need to ask multiple times or ORDER it to do work, it just gets stuff done — even beyond 50% the token limit and after chat compaction! This kind of message is a thing...
LLMs Still Far From Their Intelligence Ceiling
LLMs haven't hit their "smartness" limits yet.
Six-Layer Stack Powers Autonomous Generative AI Agents
The Generative AI ecosystem is evolving into a full tech stack — powering autonomous AI agents. From infrastructure and LLMs to RAG pipelines, agent behaviors and orchestration layers, this framework shows the 6 layers driving next-gen AI systems. Credit: @goyalshalini #AI #GenerativeAI #AgenticAI...