HN kanemarais 5/6/2026

Show HN: Extract and Run Code from Video

YouNote tool extracts and runs code from video tutorials via multimodal interface. Developer tool with novel approach to learning from video content.

HN steveharing1 5/6/2026

Ling 2.6 (Flash and 1T): Efficient Open Models Competing on Agentic Benchmarks

Ling 2.6 open models from Ant Group: 1T parameter flagship and 104B flash variant. MIT licensed, competitive on agentic benchmarks. Underrated alternative to DeepSeek/Qwen.

HN WhyNotHugo 5/6/2026

Meat-Based LLM Proxies

Opinion piece about human intermediaries relaying LLM responses. Speculative commentary without technical substance.

HN berlianta 5/6/2026

Supercomputer networking to accelerate large scale AI training

OpenAI partners with AMD, Broadcom, Intel, Microsoft, NVIDIA on MRC protocol for GPU networking in large-scale AI training. Infrastructure for frontier models.

HN manoj079 5/6/2026

81.1% vs. 13.6%: measuring retrieval accuracy for AIcoding context without embed

SigMap benchmark measures file retrieval accuracy for AI coding context. 80% hit rate vs 13.6% baseline on real repositories. Technical benchmark.

HN bundie 5/6/2026

Stop Sending IDE-Catchable AI Code Errors to Review

Analysis of AI code quality issues in review processes and JetBrains IDE integration. Addresses developer tooling challenges from AI-generated code.

HN hacker2builder 5/6/2026

Show HN: ContextWizard – AI context manager with undo and drag-drop

ContextWizard browser extension manages AI context with drag-drop bookmarks, undo, encrypted sync. Developer tool for interfacing with ChatGPT, Claude, Gemini.

HN blenderob 5/6/2026

Not your Weights, Not your Brain

Essay on exocortex concept using agents for research/synthesis/code generation, emphasizing persistent memory and open source importance.

HN nerptastic 5/6/2026

Ask HN: Is writing code by hand still a necessary skill for developers?

Discussion thread questioning whether hand-coding remains necessary as AI agents become primary development interface. Community perspective on AI and developer roles.

HN macraft 5/6/2026

Show HN: I built an app to learn anything I want during my school breaks

App for learning during school breaks using AI to generate courses on chosen topics. Consumer learning tool with AI generation.

HN mikelgan 5/6/2026

Are you addicted to your AI chatbot?

Research on AI chatbot addiction presented at CHI 2026 explores how chatbot design features may contribute to dependency behaviors.

HN spk_ 5/6/2026

Composing Claude Code Skills

Guide on composing Claude Code Skills following Unix principles, with auto-triggering skills for improved agent performance.

BL 5/6/2026

AlphaEvolve: How our Gemini-powered coding agent is scaling impact across fields

Gemini-powered coding agent for algorithm design and optimization deployed across Google infrastructure. Demonstrates real-world impact on discovery and performance.

HN lschneider 5/6/2026

Hackable PyTorch RL Library with Distributional Algorithms (D4PG, DSAC, DPPO)

Fast GPU-accelerated PyTorch reinforcement learning library implementing D4PG, DSAC, DPPO algorithms, forked from rsl_rl.

HN danicuki 5/6/2026

Automated pipeline that turns a narration recording into a YouTube video

Automated pipeline converting narration recordings to published YouTube videos using Google's Lyria 3 Pro model with idempotent steps.

HN emilypellegrini 5/6/2026

SEO: Medium Post Automation Agent

Stub entry about Medium post automation agent, no substantive content.

HN lyuata 5/6/2026

What Is an AI SDR? How AI Sales Agents Work in 2026

Overview of AI SDR (Sales Development Representative) agents, market size ($4B), capabilities in prospecting and outreach automation, and cost analysis.

HN throawayonthe 5/6/2026

We're Not Building AI Features for the Money

Zed editor maintainer defends Parallel Agents feature development as principled engineering choice, not money-driven, addresses user concerns about priorities.

HN david_d8912 5/6/2026

How to improve code quality of Claude Code and codex (on 2026-05)

Discussion on improving code quality of Claude Code and GPT-5.5 agents for software development tasks.

HN Palmik 5/6/2026

Exclusive / House Committees Probe Cursor Parent, Airbnb over Chinese AI

House committees probe Anysphere (Cursor) and Airbnb over use of Chinese AI models and geopolitical concerns.

HN ingve 5/6/2026

Running a Local LLM Coding Server on MacBook Pro M5 Pro 48 GB

Setup guide for running local LLM coding server on MacBook Pro M5 with 48GB, exposing OpenAI-compatible API without cloud.

HN mpgirro 5/6/2026

GitAgentProtocol (Open Gap)

GitAgentProtocol (Open GAP) is framework-agnostic standard for defining AI agents portably across Claude Code, OpenAI, LangChain, CrewAI, and AutoGen.

HN nikoma777 5/6/2026

We rebuilt our auth back end in under an hour using a DSL/compiler approach

Carrier is a compiled DSL/language for building API services and backend logic with automatic compiler generation of code, tests, and OpenAPI specs.

HN tommyjepsen 5/6/2026

Show HN: Cybersecurity Phishing Guard for Chrome using local LLMs for privacy

Chrome extension using local LLMs for privacy-preserving phishing detection analyzing 6 signals without data leaving the machine.

BL 5/6/2026

How ChatGPT learns about the world while protecting privacy

Guide to ChatGPT model training, privacy safeguards, and privacy controls for users.

HN bozdemir 5/6/2026

Q8 KV cache lets a 30B model fit 100K context on a 24 GB RTX 5090

Experimental setup achieving 100K context window for 30B model on 24GB RTX 5090 using Q8 KV cache quantization with Ollama.

HN simmestd 5/6/2026

Show HN: Trupositive – Tag Terraform/CFN Resources with Git Commit SHA

Trupositive automatically tags infrastructure resources with Git metadata (commit SHA, branch) for Terraform and CloudFormation.

HN populer 5/6/2026

ByteDance tests paid subscriptions in China for AI app Doubao

ByteDance introduces three-tier paid subscription plans for its Doubao AI chatbot app in China.

HN gengeo-ai 5/6/2026

GenGEO: A binary trust registry for AI agent transactions

Machine-readable trust registry for AI agents to verify merchant safety before making autonomous transactions.

HN joozio 5/6/2026

What if LLMs are mostly crystallized intelligence?

Analysis of LLM capabilities distinguishing crystallized intelligence (pattern learning from training data) from fluid intelligence (novel problem-solving), suggesting LLMs excel at trained domains but lack human-like reasoning.

HN WayLonWen 5/6/2026

SensorHub – The event-driven version of Clawhub (giving AI agents "ears")

Open registry of W2A-compatible sensors providing event-driven input streams for AI agents, with install commands.

HN shritama_saha 5/6/2026

Integrate Cashfree Payments in less than 7 minutes

Article on using AI coding assistants for payments integration, noting limitations in product-specific tasks like webhook verification.

HN babelfish 5/6/2026

The guide to RL environments: building and scaling them in the LLM era

Guide to building and scaling reinforcement learning environments in the era of LLM-based agents.

HN manupareek 5/6/2026

Generative UI Browser

Interactive experiment using generative UI to create full-page images from search queries, navigating by tapping.

HN Zombwaffle 5/6/2026

Why LLM APIs Shouldn't Ship UTF-8", "Stop Wasting Bandwidth on LLM Text APIs

Binary token-native transport protocol for LLM APIs to reduce bandwidth by keeping token IDs as integers instead of converting to UTF-8 JSON.

HN zsqzz 5/6/2026

CLI2API: Turn Your Claude Subscription into an OpenAI-Compatible API

Wrapper tool converting Claude CLI into OpenAI-compatible API with concurrency, retry logic, and prompt caching.

HN sosuke 5/6/2026

The Workflows I Created Using LLMs Became My Ecosystem

Developer built compiler to manage AI coding agent skills across multiple platforms (Claude, Cursor, Windsurf) with unified source format.

Ax Jay Lee, Hanqi Su, Marco Macchi, Adalberto Polenghi, Wei Wu, Zhiheng Zhao, George Q. Huang, Kiva Allgood, Devendra Jain, Benedikt Gieger, Vibhor Pandhare, Soumyabrata Bhattacharjee, Ram Mohril, Lingbao Kong, Qiyuan Wang, Xinlan Tang, Sungjong Kim, Chan Hee Park, Byeng D. Youn, Guo Dong Goh, Xi Huang, Wai Yee Yeong, Yung C Shin, He Zhang, Zitong Wang, Fei Tao, Jagjit Singh Srai, Satyandra K. Gupta, Byung Gun Joung, Albin John, John W. Sutherland, Sang Won Lee, Olga Fink, Vinay Sharma, Faez Ahmed, Wei Chen, Mark Fuge, Arild Waaler, Martin G. Skj{\ae}veland, Dimitris Kyritsis, Wei Chen, VispiNevile Karkaria, Yi-Ping Chen, Ying-Kuan Tsai, Joseph Cohen, Xun Huan, Jing Lin, Liangwei Zhang, Gregory W. Vogl, Aaron W. Cornelius, Xiaodong Jia, Dai-Yan Ji, Takanobu Minami, Ruoxin Wang 5/6/2026

2026 Roadmap on Artificial Intelligence and Machine Learning for Smart Manufacturing

2026 roadmap addressing AI/ML challenges in smart manufacturing including data management, system integration, and industrial deployment.

Ax Viet Trinh, Tan Nguyen, Minh-Huyen Phan, Quan Luu 5/6/2026

AI Agents for Sustainable SMEs: A Green ESG Assessment Framework

AI agent system built on n8n platform for automated ESG performance classification and assessment in European SMEs using expert-validated baselines.

Ax Gouki Minegishi, Hiroki Furuta, Takeshi Kojima, Yusuke Iwasawa, Yutaka Matsuo 5/6/2026

Understanding Emergent Misalignment via Feature Superposition Geometry

Geometric analysis of emergent misalignment in LLMs through feature superposition, explaining how fine-tuning on narrow tasks induces harmful behaviors.

Ax Navapat Nananukul, Mayank Kejriwal 5/6/2026

ClinicBot: A Guideline-Grounded Clinical Chatbot with Prioritized Evidence RAG and Verifiable Citations

Clinical chatbot using prioritized evidence RAG with guideline-grounding and verifiable citations to reduce hallucination in medical diagnosis.

Ax Basit Mahmud Shahriar, Md Habibor Rahman 5/6/2026

A Knowledge-Driven LLM-Based Decision-Support System for Explainable Defect Analysis and Mitigation Guidance in Laser Powder Bed Fusion

Knowledge-driven LLM decision-support system with ontology integration for defect diagnosis and mitigation in laser powder bed fusion manufacturing.

Ax Shakeel Sheikh, Patrick Marmaroli, MD Sahidullah, Slim Ouni, Fabrice Hirsch, Goncalo Leal, Bjorn W Schuller 5/6/2026

Virtual Speech Therapist: A Clinician-in-the-Loop AI Speech Therapy Agent for Personalized and Supervised Therapy

LLM-based intelligent agent platform for stuttering assessment and personalized therapy with clinician-in-the-loop workflow.

Ax Jinpai Zhao, Albert Cerrone, Joannes Westerink, Clint Dawson 5/6/2026

Towards Multi-Agent Autonomous Reasoning in Hydrodynamics

Multi-agent system prototype for scientific workflows in hydrodynamics, addressing context limitations of single-agent LLM systems.

Ax Jay Bhan, Nicole Nobili, Srinivasan Raghuraman, Patrick Langer 5/6/2026

New Bounds for Zarankiewicz Numbers via Reinforced LLM Evolutionary Search

LLM-guided evolutionary search establishing new exact Zarankiewicz numbers and bounds through reinforced optimization.

Ax Ravi Ranjan, Utkarsh Grover, Xiaomin Lin, Agoritsa Polyzou 5/6/2026

PERSA: Reinforcement Learning for Professor-Style Personalized Feedback with LLMs

RLHF-based approach for adapting LLMs to match instructor style in automated educational feedback while preserving diagnostic accuracy.

Ax Zephaniah Roe, Jack Sanderson, Dang Nguyen, Julian Huang, Todd Nief, Aryan Shrivastava, Chenhao Tan, Ari Holtzman 5/6/2026

Iterative Finetuning is Mostly Idempotent

Empirical study showing iterative finetuning on model outputs mostly produces idempotent behavior rather than amplifying tendencies.

Ax Sheldon Yu, Yingcheng Sun, Hanqing Guo, Julian McAuley, Qianqian Tong 5/6/2026

A Low-Latency Fraud Detection Layer for Detecting Adversarial Interaction Patterns in LLM-Powered Agents

Defense system detecting adversarial interaction patterns in LLM agents through low-latency anomaly detection on agent behavior.

Ax Tanav Singh Bajaj, Nikhil Singh, Karan Anand, Eishkaran Singh 5/6/2026

Position: Safety and Fairness in Agentic AI Depend on Interaction Topology, Not on Model Scale or Alignment

Position paper arguing multi-agent safety depends on interaction topology rather than individual model alignment or scale.

Ax Sheridan Feucht, Tal Haklay, Usha Bhalla, Daniel Wurgaft, Can Rager, Rapha\"el Sarfati, Jack Merullo, Thomas McGrath, Owen Lewis, Ekdeep Singh Lubana, Thomas Fel, Atticus Geiger 5/6/2026

Arithmetic in the Wild: Llama uses Base-10 Addition to Reason About Cyclic Concepts

Mechanistic analysis of how Llama-3.1-8B performs cyclic reasoning through base-10 addition rather than modular arithmetic.