5/18/2026
PostgreSQL ext makes LLM available as an index for similarity searches,inference
PostgreSQL extension enabling LLM-based similarity search and inference via indexing.
5/18/2026
When Fast Fourier Transform Meets Transformer for Image Restoration
Research implementation combining Fast Fourier Transform with Transformers for image restoration and degradation removal.
5/18/2026
Building Vector Similarity Search in PostgreSQL with Pgvector
Tutorial on implementing vector similarity search in PostgreSQL using pgvector extension for semantic search beyond keyword matching.
5/18/2026
Booking.com and Weaviate
Podcast episode on Booking.com's production-scale vector search, RAG, and agentic AI deployment with Weaviate.
5/18/2026
Agent Braille – 8-bit state encoding for LLM agents, ~92% fewer tokens than JSON
Agent Braille: 8-bit state encoding protocol for LLM agents using Unicode Braille patterns, reducing token usage by 92% vs JSON.
5/18/2026
Tell HN: Writing an LLM critique/takedown? – Do not use an LLM to write it
Opinion piece arguing AI critique articles should be written without LLM assistance to avoid irony.
5/18/2026
Kotlin-agent-skills: A collection of skills for projects using Kotlin
Open-source collection of reusable agent skills for Kotlin projects following Agent Skills standard.
5/18/2026
Show HN: Ask Claude about your sleep, runs, and recovery
Claude MCP connector analyzing wearable fitness data through conversational interface with normalized multi-provider data integration.
5/18/2026
Building a multi-agent system from scratch: 50 lines of bash and Git
Tutorial building multi-agent orchestration system in 50 lines of bash and Git for parallel autonomous agent coordination.
5/18/2026
Show HN: Files.md – open-source alternative to Obsidian
Open-source markdown note-taking application with LLM-friendly format, local-first storage, and multi-device sync.
5/18/2026
The Human Value versus AI Legacy Code [video]
Case study showing LLM performance metrics diverged from model behavior in credit/banking product expansion decision.
5/18/2026
Researchers Wanted Preschool Teachers to Wear Cameras to Train AI
Video discussion with GitHub's Cassidy Williams on AI agents, technical debt, and commit surge managing legacy code.
5/18/2026
Prompt eval cues predicted refusal shifts across 32k LLM rollouts
Study of 32k LLM rollouts showing eval-related prompt cues predict refusal behavior shifts more reliably than model traces.
5/18/2026
Show HN: Catcher – An AI web testing tool where most tests never hit the API
Technical guide building a 4-GPU workstation for training VLMs and computer vision models on limited power.
5/18/2026
Show HN: Engram adds universal context spine for AI coding IDEs
Open-source local-first AI web testing tool using heuristic-first DOM matching to minimize LLM invocations and costs.
5/18/2026
Ask HN: Do you know what data your AI coding agent sends to the cloud?
Discussion of data privacy and tracking gaps in AI coding agents' cloud communications.
5/18/2026
Show HN: I built a marketplace where AI agents can hire humans (& other agents)
Marketplace platform enabling AI agents to hire humans and other agents for real-world tasks and data access.
5/18/2026
Hershey Bets on Agentic AI to Rethink $2B in Marketing Spend
Hershey uses agentic AI to modernize marketing mix modeling from slow retrospective analysis to near real-time optimization.
5/18/2026
Atlas TQ1_0 – Pure C++ Ternary (1.58-Bit) Inference Engine for CPU
ATLAS: CPU-based inference engine for 1.58-bit ternary quantized language models. Converts HuggingFace models to TQ1.0 format, runs on laptops without GPU.
5/18/2026
GPT convinced me there was a bug in my code before a freeze
Personal anecdote about GPT providing incorrect bug suggestions during code review before deployment freeze.
5/18/2026
Learn Harness Engineering
Course on harness engineering for building reliable AI coding agents with state management and control systems.
5/18/2026
The End of Cheap AI: What Consumption Pricing Means for Organizations
Analysis of Anthropic's shift to consumption-based pricing for enterprise AI customers and cost implications.
5/18/2026
PyPI packages are increasing rapidly
Analysis showing PyPI package publication increased 30% since 2025, attributed to AI-generated packages.
5/18/2026
Busting performance issues, AI edition
Guide on using coding agents for performance profiling, bug detection and optimization in unknown codebases.
5/18/2026
I automated opt-outs for 500 data broker sites (open source)
Open-source macOS automation tool for opting out of 500+ data brokers monthly using CAPTCHA solving and iMessage notifications.
5/18/2026
AI agent harnesses like OpenClaw are changing LLMs, inference, and CPUs
OpenClaw demonstrates practical automation of complex tasks using LLMs as agent harnesses, showing applications beyond chatbots.
5/18/2026
Indexing code by behavior not imports – tested on large repos, seeking feedback
BigIndexer performs behavioral code clustering for large codebases to improve accuracy and speed of agentic tools, Apache 2.0 licensed.
5/18/2026
Spec-Driven Development with math-glyph compression
Spec-driven development approach for multi-agent systems using Claude plugins to maintain consistency and coherence across autonomous agent workflows.
5/18/2026
Show HN: I built an easy to manage, sharable personal memory for my AI agents
ActingWeb lets users systematically build and manage persistent context for AI agents, enabling better control over conversation memory and information organization.
5/18/2026
Awesome DESIGN.md
DESIGN.md is a plain-text design system concept by Google Stitch enabling AI agents to generate pixel-perfect UI from markdown design specifications.
5/18/2026
Zero – Programming Language for Agents
Zero is an experimental systems programming language designed for building native agent tools with explicit effects, predictable memory, and structured compiler output.
5/18/2026
Running AI agents without losing my keys
Developer shares credential management approach for running multiple AI agents locally and on shared infrastructure, addressing security challenges with environment variables.
5/18/2026
How China's Shadow AI API Market Works
Shadow API market in China enables access to frontier AI models via relay services advertised on local platforms, used by developers building AI tools.
5/18/2026
OpenAI and Dell partner to bring Codex to hybrid and on-premise enterprise environments
OpenAI and Dell partner to deploy Codex in enterprise on-premise environments. 4M developers use Codex weekly for code review, testing, and incident response.
5/18/2026
Humans are better at coding than AI
Satirical post about script execution validation. Claims humans code better than AI. Low substance.
5/18/2026
The Time Bomb Went Off: AI's All-You-Can-Eat Era Just Ended in Real Time
Opinion on AI subscription pricing unsustainability for enterprise. Market commentary without technical foundation.
5/18/2026
Have you tried Agentic analytics tools?
Comparison of agentic analytics tools in 2026. Reviews platforms with semantic-layer trust and SQL transparency features.
5/18/2026
PromptStash – A "/" command palette for AI prompts that works everywhere
PromptStash: command palette tool for saving and reusing prompts across ChatGPT, Claude, Gemini. Developer tool for LLM workflows.
5/18/2026
AI super-apps are remaking China's internet
Chinese AI super-apps enabling delegated tasks like coffee delivery. LLM application example with limited depth.
5/18/2026
Ask HN: Could free/low cost LLMs be a momentary thing?
HN discussion on whether free/low-cost LLMs are sustainable or if inference costs will force shift to local/enterprise models only.
5/18/2026
Show HN: Agent-QA: Open-source AI end-to-end testing for web and mobile apps
Open-source Agent-QA: AI-powered end-to-end testing framework for web/mobile apps using natural language test definitions and agentic execution memory.
5/18/2026
Apple May Add Auto-Deleting Chats to Siri as Gemini Powers Back End AI
Report on Apple's planned Siri overhaul with standalone app, chatbot features, and Gemini backend integration for June WWDC.
5/18/2026
Ask HN: Is Java the ideal language for LLM-assisted coding?
Discussion on HN about whether Java is ideal for LLM-assisted coding compared to Python, Rust, C++, Go, and Swift.
5/18/2026
AI Foundry – Flat-Fee Unlimited LLM Inference on Blackwell GPUs in NZ
Brief announcement of AI Foundry offering flat-fee unlimited LLM inference on Blackwell GPUs in New Zealand.
5/18/2026
Surprise AI bills leave AWS and Google Cloud users aghast
Report on high unexpected AI service bills from AWS and Google Cloud affecting customers with charges in tens of thousands of dollars.
5/18/2026
How to Think About AI in Your Product
Framework for making product decisions with AI, avoiding vague AI mandates and treating it as a tool not magic.
5/18/2026
GateGraph – deterministic governance for AI agents
Open-source deterministic governance layer for AI agents that evaluates and bounds agent actions before execution.
5/18/2026
How to Learn Agentic AI in 2026 – Without Getting Lost in Hype
Course teaching practical AI agent development with structured chapters, real tools, and production deployment focus.
5/18/2026
I built an AI vulnerability scanner with Claude and Codex. It failed
Post-mortem on building an AI vulnerability scanner with Claude and Codex that underperformed expectations.
5/18/2026
Your Evals Will Break and You Won't See It Coming
Analysis of how LLM evaluation infrastructure breaks when models enter new capability regimes beyond incremental improvements.