We are creating a superlearner that discovers all knowledge from its own experience, from elementary motor skills through to profound intellectual breakthroughs.
This superlearning capability – the ability to endlessly discover knowledge and skills, without relying on human data – will be driven by the world’s most powerful reinforcement learning algorithms.
The superlearner is expected to rediscover and then transcend the greatest inventions in human history, such as language, science, mathematics and technology.
If successful, this will represent a scientific breakthrough of comparable magnitude to Darwin: where his law explained all Life, our law will explain and build all Intelligence. — Read More
Recent Updates Page 20
Set Up Useful AI Teammates With New ChatGPT Workspace Agents
In this guide, you will learn how to set up AI teammates that are actually useful in ChatGPT’s new Workspace Agents tool. The goal is one daily agent that handles a recurring morning task for you, instead of one more prompt you have to remember to run yourself.
You will build a daily agent that owns one recurring workflow and runs it for you each morning.
In our demo, that agent reviewed a Notion database of published guides, looked for useful patterns, and generated three new guide ideas every day. — Read More
Google says 75% of the company’s new code is AI-generated
Three-quarters of new code created inside Google is now generated by AI and reviewed by human engineers, the company said Wednesday.
That number has been notching up in recent years. As of October 2024, around a quarter of the company’s code was AI-generated, Google said at the time. Last fall, it said the number had risen to 50%.
The company has been pushing employees to use AI both for coding and other tasks. Google CEO Sundar Pichai said in a blog post on Wednesday that the company was shifting to “truly agentic workflows” with its engineers running more autonomous tasks — Read More
Efficient Video Intelligence in 2026
Five years ago, video understanding mostly meant action recognition on Kinetics-400 or short-clip captioning on MSR-VTT. Today, vision-language models reason about hour-long footage, on-device tracking segments any object at 16 FPS on a phone, and a single 100M-parameter encoder can match domain experts across image understanding, dense prediction, and VLM tasks. The shift came from rethinking what a video model needs to do, and from taking deployment constraints seriously.
This post walks through where efficient video intelligence stands in April 2026, following how a video system processes its input from raw frames through spatial perception, long-form temporal understanding, multimodal fusion and reasoning, and the deployment stack that makes any of it shippable.
A note up front: the post leans heavily on research from my own group, including EUPE, the EfficientSAM / Efficient Track Anything / EdgeTAM compression line, LongVU, Tempo, EgoAVU, VideoAuto-R1, DepthLM, and ParetoQ. I have tried to place each piece against the parallel and competing work in its section, but this is a perspective from inside one research program rather than a neutral survey. — Read More
The World Can’t Keep Up With AI Labs
Late last year a new AI psychosis kicked off. This time it was coding agents.
People started saying this is a new era in programming, blah blah blah.
A few months later, we’ve got more than just claims. We’ve got numbers. And they say something unusual is happening in the market.
Coding agents are the first AI product people are paying for at volume and regularly. Because it directly speeds up their work. It’s too early to claim businesses are replacing whole processes with agents across the board. But compute demand has started growing faster than anyone can build it out.
Here’s why this moment is different, why nobody’s ready, and what I took from it personally. — Read More
Meta Is Cutting 8,000 Jobs While Posting Record Profits. That’s the New Normal.
Meta Platforms reported $201 billion in revenue in 2025. Net income was a record. Free cash flow was a record. The stock closed the year at an all-time high.
On Thursday, the company announced it is laying off 8,000 people.
Those two things are not a contradiction anymore. They are the business model. — Read More
HX Is The New UX: What You Need To Know About Harness Experience.
For thirty years, the central obsession of product design has been a single question: how do we make it easier for a human to click the right button? We built funnels. We A/B tested button colors. We agonized over empty states and loading spinners.
That era is ending — not gradually, but structurally.
… Agents don’t navigate UIs. They negotiate with systems. And when the agent is the primary “user” of software, the human behind it occupies an entirely different role — one for which we have almost no design vocabulary. Until now.
HX — Harness Experience — is the design discipline governing the interface between a human and their agentic fleet. — Read More
The New Rules of Making Money With AI Tools
The direct answer: Making money with AI in 2026 is less about using AI and more about what you build around it. The tools are commodities. The real income lives in specialization, human judgment layered on top of automation, and solving problems so specific that generic AI can’t touch them. Operators who understand this are pulling ahead. Everyone else is racing to the bottom.
The honeymoon phase is over. Remember 2023? Selling prompt packs on Gumroad, flooding Amazon KDP with AI-written books, spinning up generic chatbot wrappers, and calling it a SaaS. For a hot minute, novelty was enough. Those days are gone — not because AI got worse, but because everyone caught up.
The barrier to entry dropped to zero. Which means zero is also what most low-effort AI hustles are now worth.
But here’s the thing nobody tells you: the real money was never in the tools. It was always in the thinking behind them. The AI landscape just finally sorted itself out enough to prove it. — Read More
Can AI Attack the Cloud? Lessons From Building an Autonomous Cloud Offensive Multi-Agent System
The offensive capabilities of large language models (LLMs) have until recently existed as theoretical risks – frequently discussed at security conferences and in conceptual industry reports, but rarely discovered in practical exploits. However, in November 2025, Anthropic published a pivotal report documenting a state-sponsored espionage campaign. In this operation, AI didn’t just assist human operators – it became the operator, performing 80-90% of the campaign autonomously, at speeds that no human team could match.
This disclosure shifted the conversation from “could this happen?” to “this is happening.” But it also raised practical questions: Can AI actually operate autonomously end-to-end, or does it still require human guidance at each decision point? Where do current LLM capabilities excel, and where do they fall short compared to skilled human operators?
To answer these questions, we built a multi-agent penetration testing proof of concept (PoC), designed to empirically test autonomous AI offensive capabilities against cloud environments. — Read More
A Hundred Robots Are Running A Bio Lab
The small robot has brushed past me five times in the last hour.
It runs loops around the perimeter of the third floor of this bio lab, serving as a courier. The machine’s job is to visit workstations and keep other robots – arms bolted to lab benches – fed with whatever they need be it pipette holders, sealed plates or something in a labeled bag. The little bot is relentless and unconcerned about me or much else beyond its job. Out of the corner of my eye, I spot chairs still rotating slowly on their bases from where it clipped them on the last pass.
About a hundred robotic arms fill this room, each one positioned beside a different scientific tool. The arms must deal with centrifuges, incubators, chambers and tubes. They run simultaneously and continuously. The small robot links them together, ferrying consumables between stations the way a junior scientist carries things between benches. Except the benches are robots. And so is the assistant. — Read More
Rick's Cafe AI 