1081
New
The latest blog posts to take off on Hacker News. 1,521 posts.
1082
Claude Opus 4.7
Claude Opus 4.7 is now generally available, with notable gains over Opus 4.6 in advanced software engineering and the most difficult tasks.
1083
Qwen3.6-35B-A3B: Agentic coding power, now open to all
Qwen offers comprehensive functionality spanning chatbot, image and video understanding, image generation, document processing, web search integration, tool utilization, and artifacts.
1084
Cloudflare's AI Platform: an inference layer designed for agents
We're building AI Gateway into a unified inference layer for AI, letting developers call models from 14+ providers. New features include Workers AI binding integration and an expanded catalog with multimodal models.
1085
1086
The local LLM ecosystem doesn’t need Ollama
Ollama gained traction by being the first easy llama.cpp wrapper, then spent years dodging attribution, misleading users, and pivoting to cloud, all while riding VC money earned on someone else's engine. Here's the full history, and why the alternatives are better.
1087
AI-assisted cognition endangers human development?
Does AI-assisted cognition threaten human development? Explore the risks of AI-assisted thinking and learn strategies to use AI tools without freezing your critical thinking.
1088
The Gemini app is now on Mac
Google is bringing the Gemini app to macOS as a native desktop experience.
1089
CPUs Aren't Dead. Gemma2B Out Scored GPT-3.5 Turbo on Test That Made It Famous
Gemma 2B scored ~8.0 on MT-Bench. GPT-3.5 Turbo scored 7.94. An 87-times-smaller model on a laptop CPU, no GPU anywhere in the stack. We published the full tape — every question, every turn, every score — so anyone can verify it. We found seven failure classes. Not hallucinations. Specific patterns: arithmetic where…
1090
Gemini Robotics-ER 1.6
Gemini Robotics ER 1.6 upgrades spatial reasoning and multi-view understanding, unlocking new capabilities like instrument reading for autonomous robots.
1091
Are the costs of AI agents also rising exponentially? (2025)
There is an extremely important question about the near-future of AI that almost no-one is asking. We’ve all seen the graphs from METR showing that the length of tasks AI agents can perform has been growing exponentially over the last 7 years. While GPT-2 could only do software engineering tasks t
1092
Cybersecurity looks like proof of work now
AISI’s tests of Mythos suggest a new security economy: to harden a system you must spend more tokens finding exploits than attackers do.
1093
Turn your best AI prompts into one-click tools in Chrome
Skills in Chrome let you discover, save and remix AI workflows — and repeat them instantly.
1094
An AI Vibe Coding Horror Story
A medical professional built a patient management app using AI, exposed all patient data unprotected to the internet, and sent voice recordings to AI services without consent. Likely violating Swiss nDSG and other laws.
1095
Can Claude Fly a Plane?
I asked Claude to look up the X-Plane 12 API and try to fly a Cessna from Hainan to somewhere nearby. It kept a pilot's log. Everything below is Claude's own notes from during the flight. I occasionally had to remind it to keep the pilot log running, and had to tell it when it crashed as xplane would quickly restart…
1096
Multi-Agentic Software Development Is a Distributed Systems Problem
Recently, I've been thinking a lot about scaffolding and languages for managing systems of LLMs coordinating with each other — new programming languages might be the ideal solution for this area.
1097
Average is all you need
LLMs will make more of your average stuff. And that's OK. Here's how average SQL beats fancy attribution models when your agent writes the queries.
1098
AI could be the end of the digital wave, not the next big thing
The AI boom is more likely a marker of the end of the 50-year digital boom than the start of a new wave of innovation.
1099
Apple's accidental moat: How the "AI Loser" may end up winning
Apple’s accidental moat
1100
Tech valuations are back to pre-AI boom levels
The chart below compares the forward P/E ratios for the S&P 500 and the S&P 500 Information Technology sector. Subscribe for daily updates.
1101
1102
Why AI Sucks at Front End
How can it generate 3D worlds, videos, images and entire web pages, but still suck at front-end?
1103
1104
Exploiting the most prominent AI agent benchmarks
Every week, a new AI model climbs to the top of a benchmark leaderboard. Companies cite these numbers in press releases. Investors use them to justify valuations. Engineers use them to pick which model to deploy. The implicit promise is simple: a higher score means a more capable system.
1105
Small models also found the vulnerabilities that Mythos found
When AISLE tested Mythos's showcase vulnerabilities on small, cheap, open-weights models, most found the same bugs. Here's what that means for cyber.
1106
I still prefer MCP over skills
Why I believe the Model Context Protocol (MCP) is a better architectural pattern than the emerging “Skills” trend for LLM tool integration.
1107
Instant 1.0, a backend for AI-coded apps
Instant 1.0 is out! This essay shows a bunch of demos, to explain why we think Instant is the best backend for AI-coded apps. We also cover the architecture that makes all of it work.
1108
Research-Driven Agents: When an agent reads before it codes
Coding agents working from code alone generate shallow hypotheses. Adding a research phase — arxiv papers, competing forks, other backends — produced 5 kernel fusions that made llama.cpp CPU inference 15% faster.
1109
The Vercel plugin on Claude Code wants to read your prompts
09 Apr 2026 The Vercel Plugin on Claude Code wants to read all your prompts!
1110
The M×N problem of tool calling and open-source models
Tool calling with closed-source models is seamless. You pass a list of functions to the API, the model calls them, you get structured JSON back.