5/15/2026
AI Wellbeing – Measuring and Improving the Functional Pleasure and Pain of AIs
Research paper measuring LLM 'functional wellbeing' and creating optimized inputs to influence model behavior.
5/15/2026
Ask HN: What is shared across participants within "AI-native" environments?
Discussion of AI-native work environments, prompting strategies, agents, and MCPs shifting focus from task execution to instruction writing.
5/15/2026
AI co-mathematician: Accelerating mathematicians with agentic AI
arXiv announcement for AI co-mathematician project using agentic AI. Header text only, full content unavailable.
5/15/2026
LLMs can't read PDFs in 2026?
Hypothetical 2026 scenario documenting poor PDF reading accuracy across Claude, ChatGPT, and Gemini. Evaluates LLM limitations on structured document extraction with ground-truth metrics.
5/15/2026
GPT convinced me there was a bug in my code before a freeze
Personal anecdote about using Claude for code review. Fragmented narrative about LLM behavior and debugging workflow. Incomplete content.
5/15/2026
We Tested DeepSeek V4 Pro and Flash Against Claude Opus 4.7 and Kimi K2.6
Benchmark comparison of DeepSeek V4 Pro/Flash against Claude Opus 4.7 and Kimi K2.6 using same methodology. Open-weight models under MIT license.
5/15/2026
Codex Harness in OpenClaw
OpenClaw agent now uses OpenAI's Codex harness as default runtime for agentic work, reducing translation overhead.
5/15/2026
15% of AI agent skill files carry hardcoded credentials with DB write access
Security research finding 15% of AI agent skill files contain hardcoded database credentials with write access.
5/15/2026
Claude Code Issue that important facts were forgotten when sessions were reseted
Persistent project memory system for Claude Code across sessions using lightweight hooks to maintain structured knowledge tree.
5/15/2026
What Coding Is Starting to Lose
Personal account of using AI assistant to fix race condition bug instead of manual debugging, reflecting on changing coding practices.
5/15/2026
From a Telegram conversation to verified AI agents in 49 days
Case study documenting 49-day timeline from Telegram conversation to deployment of verified AI agents.
5/15/2026
MCP Is Not Enough
Analysis of Model Context Protocol limitations for multi-step production workflows, arguing additional orchestration layers needed beyond MCP for complex automation.
5/15/2026
Show HN: Openvid – open-source cinematic screen recorder and mockup editor
Open-source screen recorder and video editor with cinematic effects, device mockups, and 3D capabilities using Supabase backend.
5/15/2026
Databricks brings GPT-5.5 to enterprise agent workflows
Databricks integrates GPT-5.5 for enterprise agent workflows, achieving state-of-the-art on OfficeQA Pro benchmark with 50% accuracy.
5/15/2026
How data science teams use Codex
Data science teams use Codex to convert dashboards, metrics, and raw data into review-ready analysis assets with charts and caveats.
5/14/2026
Recursive Self-Improvement Delivers New SOTA Coding Performance
Machine-readable verification layer for merchant validation in AI shopping agent systems.
5/14/2026
Raindrop – Local Agent Debugger
Research on recursive self-improvement achieving state-of-the-art coding performance in language models.
5/14/2026
LLM Policy for Rust Compiler
Policy discussion on using LLMs in Rust compiler development. Minimal details provided.
5/14/2026
LLMs run on top of an OS designed for code, not weights
Technical proposal: weight paging for LLMs on memory-constrained hardware by adapting OS virtual memory concepts to model parameters, enabling 200B+ models on 16GB machines.
5/14/2026
Shareable AI Editable Visualizations
framejs.io is an open-source embeddable web app for creating editable dashboards and visualizations. Integrates with Claude, ChatGPT, or any LLM to build custom interfaces from natural language descriptions.
5/14/2026
AI to infest eight in ten premium phones within two years
Market forecast: 80% of premium smartphones will have agentic AI capabilities by 2027. General industry prediction without technical depth.
5/14/2026
ICLR 2026 – Institutional Affiliations Dataset and Analysis
Pipeline extracting institutional affiliations from 5,356 ICLR 2026 papers into dataset and treemap visualization of AI research institutions.
5/14/2026
Show HN: Parse LLM Markdown streams incrementally on the server or client
Developer tool for parsing LLM markdown streams incrementally on server or client. Addresses performance bottleneck where frontends re-parse entire markdown documents as ChatGPT/Claude responses arrive.
5/14/2026
Vouch-compatible PR-spam reduction GitHub Action workflow and CLI tool
GitHub Action and CLI tool using multi-signal scoring to detect and block spam PRs while allowing legitimate first-time contributors.
5/14/2026
Lessons from Building AI Agents for Financial Services
Technical insights from two years building AI agents for financial services. Covers challenges with accuracy, hallucinations, real-world constraints, and reliability in high-stakes domains.
5/14/2026
Parametric CAD Bench
Parametric CAD Bench: benchmark for AI agents designing parametric 3D mechanical parts. Includes open-sourced validator, Hugging Face dataset, and leaderboard at cadbench.ai.
5/14/2026
The Spec Is the Ceiling
Analysis arguing that AI agents have commoditized downstream engineering work beyond specification, requiring organizational restructuring.
5/14/2026
Show HN: Watch a neural net discover molecules by arguing with itself
Reinforcement learning agent that generates molecular SMILES strings through adversarial self-play with live visualization.
5/14/2026
Show HN: Design.md Generator
Chrome extension extracting webpage styling into DESIGN.md/SKILL.md documentation compatible with Claude, Codex, and Cursor for AI-assisted design.
5/14/2026
Claude for Legal
Claude agents and skills reference implementation for legal workflows with multiple deployment options.
5/14/2026
Show HN: Full Stack HQ – Claude.md and Agent Stack for Claude Code
IDE plugin/framework for Claude Code agents that enforces guardrails and requires approval for agent actions.
5/14/2026
The Agent Security Stack: Transport, Identity, Policy, Runtime
Framework for securing AI agents. Covers transport, identity, policy, and runtime security layers. Addresses orchestration between specialized agents with different permissions.
5/14/2026
Our Partnership with Z.ai
Puter integrates Z.ai GLM models (GLM-5.1, GLM-5, GLM-4.7) directly in browser for 80,000+ developers. No API keys or server setup required.
5/14/2026
Sea's View on the Future of Agentic Software Development with Codex
Sea Limited deploys Codex across engineering organization with 87% weekly active users; GPT-5.5 integration for agentic workflows.
5/14/2026
A²RD: Agentic Autoregressive Diffusion for Long Video Consistency
A²RD: agentic autoregressive diffusion architecture for long video synthesis using retrieve-synthesize-refine-update cycles.
5/14/2026
Dragos Documents First LLM-Assisted Strike on Water Infrastructure in Mexico
Dragos report documents first LLM-assisted cyberattack on water infrastructure in Mexico using OpenAI/Anthropic models.
5/14/2026
An AI Agent Progresses Towards Idiocy, Not Expertise
Analysis of how AI agents degrade in performance over time on projects, opposite of human expert development trajectory.
5/14/2026
Show HN: JDS – a Copilot skill suite for structuring AI coding behavior
JDS: Copilot skill suite for structuring AI coding agent behavior through discipline-enforcing skill-based workflows.
5/14/2026
A Geometric Calculator Inside a Neural Network
Research paper discovering geometric addition mechanism in Llama 3.1 8B that manipulates circular number representations.
5/14/2026
The Impact of AI-Generated Text on the Internet
Research study analyzing prevalence of AI-generated and AI-assisted text on internet, finding 35% of new websites by mid-2025.
5/14/2026
Show HN: Velda – Run jobs with serverless GPUs, without container images
Velda serverless GPU job framework eliminating containers by mirroring local dev environments to cloud.
5/14/2026
I owe my career to open-source. I'm not sure newcomers can say the same
Career reflection on open-source community contributions amid concerns about AI-generated content flooding these spaces.
5/14/2026
Ask HN: Will writing code by hand remain a part of work?
Discussion of whether hand-coding remains relevant as AI systems perform code changes faster than humans.
5/14/2026
Gemini Omni Demo Shows AI Video Getting Better at Text
Google Gemini Omni video generation model handles text better, demonstrates mathematical proof writing with legible text.
5/14/2026
Posit AI is priced for the long run
Posit AI subscription service ($20/month) for Posit Assistant and code suggestions without token subsidies.
5/14/2026
Safety‑First AI Architecture
ElviCore: Safety-first AI architecture with hardware-rooted trust and deterministic boundaries. Page loading errors.
5/14/2026
PyTorch, rewritten from scratch in pure Rust
Ferrotorch: PyTorch reimplemented in pure Rust with autodiff, neural networks, GPU support, and JIT compilation.
5/14/2026
2.3x KV Cache Compression at 32k Context – Cut VRAM Costs by 50%
Liquid Memory: KV cache compression technique reducing VRAM costs by 50% and per-token costs for long-context LLM inference.
5/14/2026
Show HN: Claude-stash – an idea queue for Claude Code
Claude-stash: Claude Code plugin for capturing and queuing ideas during development without losing context.
5/14/2026
Firnflow: Fast search over object storage (open-source turbopuffer)
Firnflow: Open-source vector and full-text search engine using tiered storage (RAM/NVMe/S3). Alternative to Turbopuffer.