181
ARC-AGI without pretraining
We solve 20% of ARC-AGI-1 with no pretraining by minimizing the description length during inference time, with the only data involved being the target puzzle itself.
182
Writing an LLM from scratch, part 13 – attention heads are dumb
A pause to take stock: realising that attention heads are simpler than I thought explained why we do the calculations we do.
183
What GPT-OSS leaks about OpenAI's training data
OpenAI recently released their open-weights model. Here we'll discuss how that inevitably leaks some information about their model training stack, and, on the way, show that GPT-5 was trained on phrases from adult websites....
184
Waymos crash less than human drivers
Human drivers keep crashing into Waymos that aren't even moving.
185
Two things LLM coding agents are still bad at
I’ve been trying to slowly ease into using LLMs for coding help again lately (after quitting ), but something always feels off -- like we’re not quite on the...
186
LLM-powered tools amplify developer capabilities rather than replacing them
How LLM-powered programming tools amplify developer capabilities rather than replace them
187
Building better AI tools
I’ve been reading this week about how humans learn, and effective ways of transferring knowledge. In addition, I’ve also had AI in the back of my mind, and...
188
LLMs are cheap
A lot of people think LLMs are expensive to operate. That was true a couple of years ago, but these people haven't updated their views after an approximately 1000x reduction in prices over two years.
189
Seven replies to the viral Apple reasoning paper and why they fall short
Also: another paper that seals the deal
190
The hidden cost of AI coding
AI coding tools boost productivity but may sacrifice the flow state and deep satisfaction developers experience when writing code by hand. What are we losing?
191
AI founders will learn the bitter lesson
Historically, general approaches always win in AI. Founders in AI application space now repeat the mistakes of AI researchers.
192
The role of developer skills in agentic coding
Notes from my colleagues on AI-assisted software delivery
193
"AI-first" is the new Return To Office
A blog about making culture. Since 1999.
194
It’s still worth blogging in the age of AI
Although it might seem that AI will make it pointless, I still think it's worth blogging.
195
Snorting the AGI with Claude Code
blog - kade killary
196
Bag of words, have mercy on us
OR: Claude will you go to prom with me?
197
Grok 4
Released last night, Grok 4 is now available via both API and a paid subscription for end-users. Update: If you ask it about controversial topics it will sometimes search X …
198
Why DeepSeek is cheap at scale but expensive to run locally
Why is DeepSeek-V3 supposedly fast and cheap to serve at scale, but too slow and expensive to run locally? Why are some AI models slow to respond but fast once…
199
Don’t let an LLM make decisions or execute business logic
Don’t let an LLM make decisions or implement business logic: they suck at that.
200
201
AI is propping up the US economy
Plus how American professors are fighting back against the AI onslaught, a backlash over AI models in Vogue, and more.
202
Less is more: Recursive reasoning with tiny networks
|| Paper | Code ||
203
It's rude to show AI output to people
Feeding slop is an act of war
204
Honda: 2 years of ml vs 1 month of prompting - heres what we learned
Recalls at major automakers cost hundreds of millions of dollars a year. It’s a huge issue. To mitigate it, our company created an analytics department solely focused on categorizing warranty claims into actionable problems.
205
Anubis Works
Xe Iaso's personal website.
206
Teaching LLMs how to solid model
It turns out that LLMs can make CAD models for simple 3D mechanical parts. And, I think they’ll be extremely good at it soon.
207
My trick for getting consistent classification from LLMs
A technique for deterministic labeling from stochastic models, with benchmarked Golang implementation.
208
Hard problems that reduce to document ranking
There are two claims I’d like to make: LLMs can be used effectively1 for listwise document ranking. Some complex problems can (surprisingly) be solved by transforming them into document ranking problems. I’ve primarily explored both of these claims in the context of using patch diffing to locate N-day…
209
Gemini 3 Pro vs. 2.5 Pro in Pokemon Crystal
From 2.5 Pro's lighthouse loop to 3 Pro's "Zombie Phoenix" strategy against Red: a full breakdown of why Gemini 3 Pro is a superior long-horizon agent.
210