AI

How machines learn, reason and generate

Neural networks, language models and the math that lets software recognise, predict and create. Explained visually, without the hype.

Explainers

Short visual breakdowns of how AI works. Open one to watch it right here.

See all 3 in the library

Latest news

The newest AI stories from the daily feed, explained in plain language.

MIT Technology ReviewSep 25, 2026

The Pentagon wants $30 million to build an AI-powered lie detector

A polygraph tracks body signals such as heart rate and breathing, on the theory that lying changes them. The Pentagon's proposed Polygraph+ program would score those signals with machine learning and add standoff sensing, which reads those signals without touching the person. Experts are skeptical, since a racing pulse can signal nerves as easily as a lie, and tech-based lie detection has a long record of failure.

Read original

The VergeOct 15, 2026

ChatGPT can peek at your Chrome tabs now

ChatGPT's Chrome extension can now look at the tabs you already have open. When you tag Chrome in a prompt, it can compare products across tabs, fill out forms, or draft emails, and the sidebar now brings your full chat history to any page. ChatGPT Work can also control tabs to send messages, fill forms or download files.

Read original

AI
TechCrunchOct 15, 2026
Meta launches an open-weight AI model that runs on Macs
Open-weight models publish their internal settings, so anyone can download and run them instead of only using the maker's app. Meta says its new Muse Spark Open, an 816 billion parameter model that is natively multimodal, is the most capable open model from the US. Running it on a Mac locally needs at least a terabyte of RAM, best on a new Mac Studio.
Read original at TechCrunch (opens in a new tab)
AI
The DecoderOct 15, 2026
Black Forest Labs' open-weights image model FLUX 3 Image Klein runs on consumer GPUs with just 8 GB of VRAM
Most image generators run in the cloud, but open weights let people run them on their own computer. Black Forest Labs released FLUX 3 Image Klein, a model with 5.6 billion parameters that outputs 4K pictures, edits them at the pixel level and can take up to ten reference pictures. A lower-precision build fits on an 8 GB graphics card, drawing an image in under nine seconds on an RTX 4070.
Read original at The Decoder (opens in a new tab)
AI
The DecoderOct 15, 2026
Nvidia's Nemotron 3.5 Super VL is an open model built to run enterprise AI agents nonstop
Nvidia released Nemotron 3.5 Super VL, an open 120 billion parameter model for enterprise agents that keep working in the background. It is a mixture of experts, so only 12 billion parameters run for each token, and it uses 4-bit compression plus draft-and-check decoding that Nvidia says speeds up text generation. Smaller Bolt versions fit on machines with 24 GB of memory.
Read original at The Decoder (opens in a new tab)
AI
The DecoderOct 15, 2026
GPT-6-Astra solves 49 of 50 physics research problems at $2.40 apiece
A group of researchers built a benchmark of 50 unpublished theoretical physics problems with verifiable answers and tested AI agents on them without hints. GPT-6-Astra, running in OpenAI's Codex environment, solved 49 of 50 at under $3 each, matching specialists. The authors say agents still lack the judgment, or taste, behind real breakthroughs.
Read original at The Decoder (opens in a new tab)
AI
TechCrunchOct 15, 2026
AI agents are coming to Rivian’s cars
Rivian's cars already have an AI assistant, and its CEO says agents built by other companies, such as Anthropic, will come next. The vehicle routes each spoken request through an in-house framework that decides whether a model running on the car or one in the cloud should answer, like a traffic light. Agents cannot brake or steer; those controls stay separate.
Read original at TechCrunch (opens in a new tab)
AI
TechCrunchOct 15, 2026
Ricursive Intelligence is betting on AI-designed chips to help solve the compute crisis
Chip designers usually split a chip into small blocks, a process that takes months, so engineers build only a few kinds. Ricursive Intelligence, from the creators of AlphaChip, is building AI that lays out entire chips and improves with each attempt; one early layout was an odd donut shape that worked. They say faster design could ease the compute crunch and make efficient custom chips practical for far more uses.
Read original at TechCrunch (opens in a new tab)

All AI news

AI

Explainers

How AI chatbots write word by word

How semantic caching reuses AI answers

APIs vs MCP: how AI connects to tools

Latest news

The Pentagon wants $30 million to build an AI-powered lie detector

ChatGPT can peek at your Chrome tabs now

Meta launches an open-weight AI model that runs on Macs

Black Forest Labs' open-weights image model FLUX 3 Image Klein runs on consumer GPUs with just 8 GB of VRAM

Nvidia's Nemotron 3.5 Super VL is an open model built to run enterprise AI agents nonstop

GPT-6-Astra solves 49 of 50 physics research problems at $2.40 apiece

AI agents are coming to Rivian’s cars

Ricursive Intelligence is betting on AI-designed chips to help solve the compute crisis

How AI chatbots write word by word

How semantic caching reuses AI answers

APIs vs MCP: how AI connects to tools

The Pentagon wants $30 million to build an AI-powered lie detector

ChatGPT can peek at your Chrome tabs now

Meta launches an open-weight AI model that runs on Macs

Black Forest Labs' open-weights image model FLUX 3 Image Klein runs on consumer GPUs with just 8 GB of VRAM

Nvidia's Nemotron 3.5 Super VL is an open model built to run enterprise AI agents nonstop

GPT-6-Astra solves 49 of 50 physics research problems at $2.40 apiece

AI agents are coming to Rivian’s cars

Ricursive Intelligence is betting on AI-designed chips to help solve the compute crisis