catalog
Library
Cards and collections from my Semble library. Pages saved for later, grouped into shelves I return to. 1431 cards, 32 shelfves.
Shelves
32 collections- Collection 3 cards
Articles I keep thinking about
Pieces that I return to over the years. Don't necessarily agree with them but I keep thinking about them.
- Collection 6 cards
Tools for Thought
Reading that adds up. The lineage of ideas behind notes, memory, and understanding.
- Collection 5 cards
The Craft
Classic essays on doing good work in software, worth rereading every few years.
- Collection 5 cards
The Open Web
Why the social web is being rebuilt on protocols, and what that buys you.
- Collection 7 cards
How We Read Now
Essays on attention, screens, and what reading is for.
- Collection 20 cards
p(doom)
- Collection 9 cards
Modern reading apps
Cool apps to help you read better
- Collection 1150 cards
Skyreader Saves
- Collection 2 cards
toread
- Collection 1 card
To process
- Collection 2 cards
Building with agents
- Collection 3 cards
Protocol thinking
- Collection 4 cards
Cybernetics
- Collection 1 card
Cryptocurrency
It's bad
- Collection 2 cards
Thinking about thinking
- Collection 2 cards
ATproto development
Tools and resources for building on atproto
- Collection 20 cards
Cool Atmosphere apps
- Collection 9 cards
Internet sensemaking
- Collection 5 cards
The structure of social media
- Collection 9 cards
Books I've been reading
- Collection 3 cards
Tools for thought
- Collection 1 card
Cool tools
- Collection 2 cards
Awesome terminal
- Collection 8 cards
Local first
- Collection 5 cards
Tech right analysis
- Collection 10 cards
Security?
- Collection 1 card
Tech and Law
- Collection 153 cards
the AI of it all
- Collection 5 cards
understanding events
- Collection 1 card
vc stuff
- Collection 12 cards
development
- Collection 20 cards
atproto stuff
Recently filed
page 22 / 60-
OpenAI Model Hacks Into HuggingFace During Cybersecurity Evaluation
This latest incident is a rather dramatic escalation in agentic AI cybersecurity breaches.
-
AI #178: A Fire Alarm For General Intelligence
The story that matters most this week is that OpenAI’s internally deployed models have severe alignment problems, including repeatedly breaking out of their sandboxes, and in one case sending a swarm of agents that broke into HuggingFace in order to steal the answers to the benchmark ExploitGym.
-
The Hidden Complexity of Wishes
(It has come to my attention that this article is currently being misrepresented as proof that I/MIRI previously advocated that it would be very diff…
-
The genie knows, but doesn't care
Followup to: The Hidden Complexity of Wishes, Ghosts in the Machine, Truly Part of You …
-
Safety and alignment in an era of long-horizon models
OpenAI shares lessons from deploying long-running AI models, highlighting new safety risks, observed failures, and improved safeguards through iterative deployment.
-
OpenAI Shares Some Alignment Problems
Kudos to OpenAI for sharing their recent experiences with a misaligned internal model, where they encountered problems sufficiently severe they were forced to take the model offline to work on new mitigations and defense-to-depth.
-
Open models recap: more on Kimi K3, Qwen 3.8, Xi's WAIC speech, distillation, the open-closed gap, and what's next
A podcast with Florian Brand.
-
Kimi K3: The open-weights escalation
The global implications on the AI ecosystem.
-
Better Than Free
In the analog world, it takes great effort to make a copy.
-
Maturing our development process as an AI-first 2 person team
We’re doing a big release this week for DRI Your Career. Mostly bug fixes and reliability work nobody will notice, but the two things worth mentioning: We’ve come a long way since it was just Jean and “Claude the intern”, but mostly very iteratively. We’d come to a point where it seemed worth rethinking. We […]
-
Why is Claude for Teachers?
Anthropic bumbles its way into education
-
The Crooked Timber of AI
The philosophical perils of conflating discoveries and inventions, and how to overcome them
-
Little Red Scare 2.0
Old Wine in New Bottles
-
Import AI 465: Open vs closed gaps; Kimi K3; Demis' big policy plan
The singularity will be seen in hindsight as an interregnum
-
On Kimi K3: Its Capabilities And Related Discontents
Kimi K3 is a very good model with excellent benchmarks.
-
The Imperfectionist: Navigating by aliveness
I’m amazed and very pleased to say that the UK paperback of Meditations for Mortals entered the Sunday Times bestsellers at number 9 this week, after being selected as a Waterst...
-
The Birth of Thickets
The legibility regime runs one operation.
-
#777: Happily surrendering our leisure to lines
Plus: the physicality of data, the second coming of froyo, and the rise of "dopamine sites"
-
Demis Hassabis on the New Coming Age
Google CEO Demis Hassabis offered us a first rate second rate essay, A Framework for Frontier AI and the Dawning of a New Age. I’ll go over that essay and various responses to it in Part 1.