511
Tversky Neural Networks
Psychologically Plausible Deep Learning with Differentiable Tversky Similarity
New
The latest blog posts to take off on Hacker News. 1,227 posts.
512
Best Practices for Building Agentic AI Systems
Patterns that keep AI agent systems from falling apart in production: two-tier architecture, stateless design, orchestration, and lessons from UserJot.
513
Claude Opus 4 and 4.1 can now end a rare subset of conversations
Claude Opus 4 and 4.1 can now end a rare subset of conversations. An update on our exploratory research on model welfare.
514
Imagen 4 is now generally available
Discover Imagen 4 Fast, Google's new speed-optimized text-to-image model, now generally available with Imagen 4 and 4 Ultra in the Gemini API.
515
Using AI to secure AI
One of Anthropic's quieter releases recently was their "Security Review," where Claude Code can identify and fix security issues in your code. But how good is it really? In my case, will it find issues with code it helped me write for my newsletter service…
516
Model intelligence is no longer the constraint for automation
Why automation lags even as models get better
517
Gemma 3 270M: Compact model for hyper-efficient AI
Explore Gemma 3 270M, a compact, energy-efficient AI model for task-specific fine-tuning, offering strong instruction-following and production-ready quantization.
518
Is chain-of-thought AI reasoning a mirage?
Reading research papers and articles about chain-of-thought reasoning makes me frustrated. There are many interesting questions to ask about chain-of-thought…
519
Why LLMs can't really build software
From the Zed Blog: Writing code is only one part of effective software engineering.
520
NSF and Nvidia award Ai2 $152M to support building an open AI ecosystem
Ai2 has been awarded a combined $152 million from the U.S. National Science Foundation (NSF) and NVIDIA as part of a jointly funded project to advance our research and develop truly open AI models and solutions that will accelerate scientific discovery.
521
522
LLMs tell bad jokes because they avoid surprises
LLMs suck at comedy, art, journalism, research, and science for the same fundamental reason
523
Claude Sonnet 4 now supports 1M tokens of context
Claude Sonnet 4 now supports up to 1 million tokens of context—a 5x increase that lets you process entire codebases, synthesize extensive document sets, and build agents that maintain coherence across hundreds of tool calls.
524
Evaluating LLMs playing text adventures
When we first set up the llm such that it could play text adventures , we noted that none of the models we tried to use with it were any good at it. We dreamed of a way to compare them, but all I could think of was setting a goal far into the game and seeing how long it takes them to get there. I just realised there’s…
525
ARM adds neural accelerators to GPUs
Explore Arm Neural Technology, the first to integrate neural accelerators into GPUs, enhancing mobile graphics with AI.
526
What's the strongest AI model you can train on a laptop in five minutes?
What’s the strongest model I can train on my MacBook Pro in five minutes? I’ll give the answer upfront: the best 5-minute model I could train was a ~1.8M-param…
527
Qodo CLI agent scores 71.2% on SWE-bench Verified
Read about Qodo scores 71.2% on SWE-bench Verified in our blog.
528
529
Claude is the drug, Cursor is the dealer
Coding Tools Are Just One Front in AI’s Supply Chain Wars
530
Claude Code is all you need
I installed Claude Code in June. I'd tried Cursor and Cline and Zed and a few others, but all of them felt clunky to me because I'm used to doing nearly everything in vanilla vim and my terminal. Claude Code was the first tool I tried that felt like it fit into my workflows perfectly rather than needing me to adapt to…
531
Flock Now Using AI to Report to Police If Our Movement Patterns Are "Suspicious"
Company crosses a dangerous line by beginning to offer AI suspicion-generation functions
532
I tried coding with AI, I became lazy and stupid
Robots won't replace me, but another human probably will.
533
GPT-OSS vs. Qwen3 and a detailed look how things evolved since GPT-2
And How They Stack Up Against Qwen3
534
MCP: An (Accidentally) Universal Plugin System
Or: The Day My Toaster Started Taking Phone Calls
535
LLMs aren't world models
I believe that language models aren’t world models. It’s a weak claim — I’m not saying they’re useless, or that we’re done milking them. It’s also a fuzzy-sounding claim — with its trillion weights, who can prove that there’s something an LLM isn't a model of? But I hope to make my claim clear and persuasive enough…
536
GPT-5: Overdue, overhyped and underwhelming. And that's not the worst of it
A new release botched … and a breaking research new paper that spells trouble
537
GPTs and Feeling Left Behind
Every time that I read some blog post about “coding with AI”, or how cool new models write entire libraries by themselves, I feel like I’m lagging behind, like I’m missing out on some big, useful tool, and my skills are about to become obsolete very soon.
538
How I code with AI on a budget/free
A personal guide on using AI Code Prep GUI, Cline, and various free AI web interfaces like Gemini, Grok, Deepseek, and Poe for cost-effective coding and debugging.
539
The current state of LLM-driven development
I spent the past ~4 weeks trying out all the new and fancy AI tools for software development.
540
My Lethal Trifecta talk at the Bay Area AI Security Meetup
I gave a talk on Wednesday at the Bay Area AI Security Meetup about prompt injection, the lethal trifecta and the challenges of securing systems that use MCP. It wasn’t …