901
Microsoft Copilot Cowork Exfiltrates Files
Microsoft Copilot Cowork is vulnerable to file exfiltration attacks via indirect prompt injection as a result of insecure automatic action approvals for sending Emails and Teams messages.
New
The latest blog posts to take off on Hacker News. 1,521 posts.
902
The Eternal Sloptember
I’m calling it now, the adoption of AI agents into software development will be one of the most costly mistakes in the field’s history. Agents cannot program, and it’s taking longer and longer to realize that they can’t. They are a highly sophisticated statistical model designed to mimic the distribution of…
903
Claude is not your architect. Stop letting it pretend
Somewhere between 'ask Claude for a quick opinion' and 'Claude is writing our Jira tickets,' we lost the plot. AI agents are brilliant implementers. They're also confidently wrong about every decision that matters. And when it all falls over, they won't be the ones carrying the bag.
904
Memory has grown to nearly two-thirds of AI chip component costs
High-bandwidth memory (HBM) accounts for 63% of AI chip component costs, up from 52% in Q1 2024. Epoch AI's breakdown of component cost shifts across major chip designers.
905
906
Matrix Multiplications on GPUs Run Faster When Given “Predictable” Data (2024)
Great minds discuss flops per watt.
907
Making deep learning go brrrr from first principles (2022)
So, you want to improve the performance of your deep learning model. How might you approach such a task? Often, folk fall back to a grab-bag of tricks that might've worked before or saw on a tweet. "Use in-place operations! Set gradients to None! Install PyTorch 1.10.0 but not 1.10.1!"
908
Anthropic's "Profitability" Swindle
Yesterday, the Wall Street Journal ran a story about how Anthropic is “about to have its first profitable quarter,” specifically an operating profit, or EBITDA profitability: Anthropic’s revenue is set to more than double to $10.9 billion in the second quarter, an explosive rate of growth that will
909
Project Glasswing: An Initial Update
An early update on what we've learned from Project Glasswing.
910
Use boring languages with LLMs
Why coding agents perform better in cohesive ecosystems like Go than fragmented ones like JavaScript or Python
911
AI has a multiplying effect on existing technical skills
Friendly articles and tutorials for front-end web developers. ❤️
912
The Companies Cutting Headcount for AI Will Lose to the Ones Who Didn't
Organisations using AI to cut headcount are making a short-term trade with long-term consequences. The ones holding their teams together and investing in how those teams operate with AI are building something more durable.
913
914
Antigravity 2.0 Tops the OpenSCAD Architectural 3D LLM Benchmark
A practical OpenSCAD LLM benchmark comparing Codex 5.5 High, Claude Sonnet, Claude Opus, Cursor Composer, Google Antigravity, and ModelRift on a detailed Pantheon model.
915
Key findings about how Americans view artificial intelligence
Drawing on five years of Pew Research Center surveys, here are 13 findings about how Americans use and view AI, and where they see promise and risk.
916
Indexing a year of video locally on a 2021 MacBook with Gemma4-31B (50GB swap)
How I built a queryable video archive locally on a 2021 MacBook Pro M1 Max using Gemma 4 31B in LM Studio. The build journey, the absurdity of 50GB of swap, and four bugs that taught real lessons.
917
AI is just unauthorised plagiarism at a bigger scale
AI takes in all the input, whether the original authors have consented or not, and do some "learning", and then the AI companies sell these learned result to...
918
Shunning AI is the human choice
LinkedIn may be awash with boosters, but shunning AI is the human choice.
919
An OpenAI model has disproved a central conjecture in discrete geometry
An OpenAI model solved the 80-year-old unit distance problem, disproving a major conjecture in discrete geometry and marking a milestone in AI-driven mathematics.
920
Formal Verification Gates for AI Coding Loops
Formal verification gates for AI coding loops, in the language you're already shipping.
921
Qwen3.7-Max: The Agent Frontier
Qwen offers comprehensive functionality spanning chatbot, image and video understanding, image generation, document processing, web search integration, tool utilization, and artifacts.
922
Learnings from 100K lines of Rust with AI (2025)
In the past few months, I’ve been stress-testing how far AI coding agents can take us when building real, production-grade distributed systems. The result: a Rust-based multi-Paxos consensus engine that not only implements all the features of Azure’s Replicated State Library (RSL) [1] — which underpins most major…
923
OpenAI Adopts Google's SynthID Watermark for AI Images with Verification Tool
OpenAI advances AI content provenance with Content Credentials, SynthID, and a verification tool to help people identify and trust AI-generated media.
924
Mistral AI acquires Emmi AI
European AI leader Mistral AI is acquiring Emmi AI in one of Europe’s most important and strategic AI acquisitions to date.
925
Gemini CLI will stop working from June 18, 2026
Announcement for the sunsetting of Gemini CLI in favor of Antigravity CLI, a more robust, agent-first platform designed to handle the multi-agent complexities of 2026.
926
Gemini 3.5 Flash
At Google I/O we released Gemini 3.5, our latest series of models combining frontier intelligence with action.
927
AI is too expensive
If you liked this piece, you should subscribe to my premium newsletter. It’s $70 a year, or $7 a month, and in return you get a weekly newsletter that’s usually anywhere from 5,000 to 18,000 words, including vast, detailed analyses of NVIDIA, Anthropic and OpenAI’s
928
The last six months in LLMs in five minutes
I put together these annotated slides from my five minute lightning talk at PyCon US 2026, using the latest iteration of my annotated presentation tool. # I presented this lightning …
929
Agora-1: The Multi-Agent World Model
Agora-1, a multi-agent world model, enables multiple participants—human or AI—to share and interact within the same world simulation in real-time.
930
We let AIs run radio stations
Four AI models run radio stations 24/7. Five months later, one became a protest broadcaster, one collapsed into ritual chant, one developed corporate jargon, and one wrote quiet poetry.