AI Briefing
Hacker News news
The latest AI and developer news from Hacker News, with short summaries.
Visit website →EU to Mandate Labels on Realistic AI Content Starting August 2
The EU will require labels on AI-generated content that looks realistic.
Flint: A Visualization Language for the AI Era
Microsoft Research unveiled Flint, a chart-generation language for AI agents.
Predictive Speculative KV Replication for Bursty LLM Inference
BTB predicts repeated-prefix bursts and proactively replicates KV caches into GPU HBM.
Everyone is building an LLM router, but we killed ours
Manifest deprecated its LLM router based on the experience of 7,000 cloud users.
qm – A Multiplayer Agent Harness for Work
qm has released an open-source agent harness that enables collaboration on Slack and the web.
Why We Write Our Own C and C++ Inference Engines
LocalAI has revealed the performance and cost of its own C/C++ inference engines, which reduce Python dependencies.
Comparing 13 models and 4 agents on SWE Tasks: Go, Java, Python, Rust, TS
On 111 SWE problems, Fable 5 recorded the highest resolve rate at 64.5%.
Running Kimi K3 at 0.50 tok/s with 29GB RAM
WASTE ran the 2.78T Kimi K3 via disk streaming on a 64GB MacBook Pro.
Google pulled its Earth AI image generator after just one day
Google rolled back Google Earth's AI image generation feature over concerns about policy violations.
Two papers including fake authors reported to have been accepted as Oral
The academic paper review process has revealed that the problem of LLM-generated fabricated citations and fake authors has reached a serious level.
Show HN: We confirmed that distilling DeepSeek into GPT-OSS does not transfer censorship - try it yourself
A model trained on data from a Chinese model acquires the intelligence but does not inherit its political censorship.
OpenJDK's Interim Policy on the Use of Generative AI
The OpenJDK community has announced an interim policy banning contributions of content created by generative AI in order to protect code quality and intellectual property.
A Go LLM SDK for Streaming and Tool-Calling AI Backends (with a React Frontend Library)
Grafana has released a Go-based LLM SDK that is compatible with the Vercel AI SDK.
GCC Steering Committee Announces AI Contribution Policy
GCC has announced a new AI policy restricting code contributions of LLM-generated content.
ChatGPT and Roblox Set to Be Included Among EU's Strictest Platform Regulations
ChatGPT and Roblox are expected to be designated as VLOPs, the category subject to the EU's strictest platform regulations.
Ignoring the Interface Cannot Solve the Computer Use Problem
It points out that AI agents bypassing APIs instead of directly manipulating the GUI is a bottleneck for performance.
Agent-Manager: A Tmux TUI for Claude Code, Codex, and OpenCode
It is a Tmux-based TUI tool that lets you unify management and monitoring of sessions across multiple AI coding agents.
Kuna: Building a Decompiler for the Coding Agent Era
An experimental decompiler called Kuna was unveiled that uses LLMs to autonomously improve its own performance.
Launch HN: Tokenless (YC S26) – Automatic Model Switching to Cut Costs
A router service has launched that reduces inference costs by running multiple models simultaneously and quickly selecting the optimal one.
Google Disbands Nobel Prize-Winning AlphaFold Team
Google has disbanded the AlphaFold team and reassigned its personnel as part of a strategic shift centered on Gemini.