7 Oct 20265 min
SocietyNote
Why are AI agents so cute?
AI is getting better at acting on our behalf. Its little mascots look as if they would rather ask for a hug. There is a long history behind that choice.
The AI Garden newsletter
Concepts behind AI products and research
29 Sept 20265 min
SystemsNote
How can an AI do your shopping?
When you ask an AI to buy something, who clicks the buttons? And whose computer is it using?
24 Sept 20263 min
ModelsNote
How Attention Gives an LLM Its Context
An LLM reads words in the company of other words. Attention is the step that lets each piece of text draw on the pieces around it.
23 Sept 20264 min
SystemsNote
CPU, GPU and NPU: What Each One Is For
These processors can all perform calculations, but they are built for different kinds of work. The useful question is not which one is fastest. It is which one fits the task.
23 Sept 20264 min
ScienceNote
System 1 and System 2: the mind’s two speeds
Jev borrowed an old psychology metaphor. The useful idea is not that the brain has two boxes, but that thinking can be automatic or effortful.
23 Sept 20265 min
ApplicationsNote
Muse wants to finish the errand, not just explain it
Meta's new personal agent can browse, fill forms, use connected apps and make purchases with approval. Its real difference is how those abilities are packaged around one person's life.
19 Sept 20265 min
ModelsNote
Jev: the AI model that answers with numbers, not words
TypeSafe AI’s first model never writes a sentence. It reads your text, answers fixed questions with probabilities, and leaves the decision to your code.
17 Sept 20262 min
ModelsNote
“Take a deep breath” helped an AI with maths
Another AI found the odd instruction. What happened in the test is more interesting than the idea of a robot calming its nerves.
16 Sept 20265 min
ModelsNote
Hugging Face: from a teen chat app to a $12.9 billion emoji
In 2017 it was an app for bored teenagers, named after the 🤗 emoji. Then its makers gave away the code behind it, and the whole AI world moved in.
16 Sept 20266 min
ModelsNote
Six ideas behind AI language models
A short history of how neural networks carry context, turn one sequence into another, and handle more work.
12 Sept 20265 min
SafetyNote
Zero-days: the software flaws nobody knows about yet
A zero-day is a weak spot in software that its makers haven’t found, so they have had zero days to fix it. AI models are now finding them too.
12 Sept 20264 min
ModelsNote
Nano Banana: the AI image model named at 2:30 a.m.
Google’s hit picture-making AI got its silly name when a product manager mashed two nicknames together in the middle of the night. Here’s that story, and how AI turns words into a picture.
12 Sept 20265 min
ModelsNote
Six of seven chatbots picked the same “random” number
Ask for a number from 1 to 50 and 27 keeps coming back. What that says about the way a model chooses every word.
11 Sept 20265 min
ApplicationsNote
Airline Ticket Prices: How AI Picks Prices in Between
Since the 1980s, airlines have sold seats at a few fixed prices and switched between them as a plane filled up. Now AI can pick a price in between and change it during the day.
9 Sept 202610 min
ScienceNote
Navier-Stokes: The million-dollar puzzle about flowing water
Start with a swirl of water, draw its motion with arrows, and use simple arithmetic to see what it would mean for the equations to break down.
9 Sept 20265 min
SafetyNote
SynthID: How a Watermark Hides Inside AI Word Choices
Google DeepMind published the method in Nature in October 2024 and ran it on about 20 million Gemini replies without anyone noticing. Claude is adopting it. This is how a watermark hides in the choice of words.
8 Sept 20265 min
SystemsNote
Quantisation: How Large Models Fit on Your Laptop
A language model is really just a very long list of numbers. Round those numbers off, and the file gets dramatically smaller and runs faster, for a small trade in quality.
8 Sept 20264 min
ModelsNote
Mixture of Experts: The Experts Aren’t Separate AIs
Inside one model, a router sends each token through a few internal blocks. Follow one token through the machinery and see what “expert” really means.
8 Sept 20264 min
ModelsNote
Long Context: Why AI Can Miss the Page That Matters
A document can fit inside an AI model’s context window and still be used unevenly. The position of the useful passage can change the answer.
8 Sept 20265 min
SystemsNote
Embeddings: How AI Search Works Without Exact Keywords
Type "money back" and a keyword search can miss the refund page. AI search turns every sentence into a point on a map and picks the nearest ones, so it finds the page. It can still pick the wrong passage first.
8 Sept 20264 min
SystemsNote
AI Memory: Why “Remember This” Is Not Model Training
An assistant can carry your preferences into a later conversation without rewriting what its language model learned. Here is where that memory actually lives.
