2026-05-15
Possible Samsung strike puts even more pressure on memory pricing
As a senior policymaker ponders whether all South Koreans should enjoy an ‘AI dividend’
Agent-Sin
AI agent that handles repeated tasks through reusable skills Discussion | Link
Some observations about NLA explanations
I used the Gemma 3 12B activation verbalizer (maps activations to English) and reconstructor (maps English to activations) described in the Natural Language Autoencoders (NLA) paper to generate a bunch...
Agentspan
Open-source runtime for durable AI agents Discussion | Link
Mix Your Own ECN-2 Development Chemicals
After the digital camera rose to prominence, it became a cool hobby to keep taking photos on film. It was even cooler if you did the same with an old...
The TPM Who Pulls Her Own Data
How the TPM role is being redefined with AI, and how a TPM can do data analysis
Operator: A look under the hood
Operator is a powerful Agent that helps you manage, optimize, and continuously improve your AI-first support organization. Here's more insight into how we built it.
Meet Operator: An Agent for your customer operations
Learn how Operator can help you understand, manage, and improve your customer experience in a way that has never been possible before.
I Built an AI That Watches My Screen and Tells Me If I Did It Right
You wish someone could sit on your shoulder, watch what you did, hold up the original instructions, and tell you exactly what matched and what didn’t. That’s what we’re building....
The hard core of alignment (is robustifying RL)
Most technical AI safety work that I read seems to miss the mark, failing to make any progress on the hard part of the problem. I think this is a...
‘Musk v. Altman’ Closing Arguments
Elizabeth Lopatto, reporting for The Verge (gift link): Today was closing arguments in the Musk v. Altman trial, and I almost feel bad writing about the unbelievable demolition derby I...
Let’s Run a Neologism Poll
After posting the previous item referencing dickpanels, a term I’ve been using since 2022, it occurred to me that they could also be called dickovers (like popovers, but dickheaded). The...
datasette-llm-limits 0.1a0
Release: datasette-llm-limits 0.1a0 This plugin works in conjunction with datasette-llm and datasette-llm-accountant to let you configure a per-user (or global) spending limit for LLM usage inside of Datasette. Configuration looks...
Convergent Abstraction Hypothesis
Tl;drConvergent abstraction hypothesis posits abstractions are often convergent in the sense of convergent evolution: different cognitive systems converge on the same abstraction, when facing similar selection pressures and learning in...
The Youth AI Safety Institute Has Margrethe Vestager’s Backing
Una Hajdari, reporting for Euronews: A new independent institute dedicated to making artificial intelligence safer for children will beformally [sic] presented at the Danish Parliament on Tuesday, with former European...
datasette-agent 0.1a2
Release: datasette-agent 0.1a2 Tool availability can now be attached to a required_permission. The default background agent tools now require the new datasette-agent-background permission. #10 Tags: datasette, datasette-agent
What Matters in Production RAG
Most of us build RAG the same way: follow a tutorial that embeds a handful of PDFs, stores the vectors in a local Chroma instance, and chains everything together with...
Aided by Mythos Preview, Researchers Announce MacOS Kernel Exploit Circumventing M5 Memory Integrity Enforcement
Calif, a security research team, on their blog: Many security experts consider Apple devices to be the most secure consumer platform. The latest flagship example is MIE (Memory Integrity Enforcement),...
Emma Baker on ADHD
A good article on ADHD by Emma Baker of Tree Number Three from Inkhaven. ADHD can be uniquely debilitating and also complicated to effectively medicate, and this article changed my...
Frictionless Verification of AI Claims
"Sometimes the AI just makes stuff up" is a problem I don't really expect to go away. In the nearterm, AI is going to keep occasionally hallucinating, or misinterpreting information....
Welcome to Postreads
Discover and follow the best content from across the web, all in one place. Create an account to start building your personalized feed today and never miss out on great reads.
Tiny Generals
Asynchronous hex-based strategy. Command armies, capture bases, and outthink opponents at your own pace.
Play NowSupport Postreads
Enjoying the service? Help me keep it running and improve it further by buying me a coffee!
Buy me a coffeeContent Timeline
81,024 articles since 2008
Trending Now
Top 5 clicked items this week out of 975 added this week
Freshly added
New feeds to discover
