2/15/2026
The Ghost in the Batch: How vLLM Silently Switches Algorithms
Analysis of vLLM's silent algorithm switching during batch inference, showing how batching changes model outputs at precision level with propagating token divergence.
2/15/2026
SurfaceDocs Now Has Versioning, Search, and Real Docs
SurfaceDocs output layer for AI pipelines enables agents to generate versioned, searchable documents with shareable URLs via Python SDK or REST API.
2/15/2026
I built a memory system that outperforms standard RAG on temporal queries -- try the live playground
Three: temporal-aware retrieval system that improves on standard RAG by tracking recency and frequency, surfacing recent information more easily than outdated data.
2/15/2026
Show HN: Governor: Let's make vibe coding safe
Tool for safe exploratory coding with minimal details provided.
2/15/2026
What If Moore Was Right?
Speculative essay on exponential AI growth and whether computational scaling laws will eventually lead to AGI-like capabilities.
2/15/2026
Former Karaoke Company Drags Logistics into the 'AI Scare Trade'
News article about logistics and AI hype/fearmongering.
2/15/2026
Show HN: Kubernetes for AI Agents
Kubernetes-based distributed management system for deploying and orchestrating multiple AI agents.
2/15/2026
UIUC 2002 – we wrote a space shooter in x86 asm. In 2026 Claude resurrected it
Claude AI resurrected a 2002 x86 assembly space shooter game, demonstrating code comprehension and resurrection capabilities.
2/15/2026
Show HN: An open-source extension to chat with your bookmarks using local LLMs
Browser extension for querying bookmarked content using local LLMs with WebLLM, no server-side processing.
2/15/2026
Claude Code at Trail of Bits
Brief mention of Claude Code at Trail of Bits with no substantive content.
2/15/2026
Generative and Agentic AI Shift Concern from Technical Debt to Cognitive Debt
Analysis of how agentic AI systems shift engineering concerns from technical debt to cognitive debt.
2/15/2026
Show HN: Alive-analysis – Git-tracked analysis notes for AI agents
Open-source workflow framework for AI coding agents enabling structured, reproducible data analysis with git-tracked markdown outputs.
2/15/2026
Dolt is a version controlled database. Dolt is Git for Data
SQL database with Git-like version control for tables, supporting fork/clone/merge operations and MySQL compatibility.
2/15/2026
AGentShield – Open benchmark of 6 AI agent security tools (537 test cases)
Open benchmark evaluating 6 AI agent security tools across 537 test cases, measuring robustness and safety.
2/15/2026
Spotify brings AI-powered Prompted Playlists to the U.S. and Canada
Spotify introduces AI-powered playlist generation feature to US and Canadian users.
2/15/2026
Adafruit – Our First Gemini Deep Think LLM-Assisted Hardware Design
Headline-only post on using Gemini LLM for hardware design at Adafruit. Insufficient technical depth.
2/15/2026
Show HN: ShadowStrike – building an open-source EDR from scratch
ShadowStrike: experimental open-source endpoint detection and response engine in pre-alpha, written in C/C++ and x86-64 assembly.
2/15/2026
Cloudflare turns websites into faster food for AI agents
Cloudflare service optimization for web scraping and content consumption by AI agents.
2/15/2026
LLM-written short story about being a LLM
Headline-only post about creative fiction generated by LLMs. Not technical or developer-focused.
2/15/2026
Tell HN: Google AI Studio docs encourage Google-discoverable open wallets
Criticism of Google AI Studio documentation regarding wallet privacy practices.
2/15/2026
The Gatekeeping Panic: What AI Actually Threatens in Software Development
Opinion piece on AI gatekeeping panic in software development, arguing real concerns are AI detection accuracy and attribution rather than tool use.
2/15/2026
Generative and Agentic AI Shift Concern from Tech Debt to Cognitive Debt
Analysis of how agentic AI systems shift engineering concerns from technical debt to cognitive debt.
2/15/2026
Ask HN: Crazy to Pivot PM into Engineering in '26?
Career advice post about transitioning from product management to software engineering.
2/15/2026
Why AI Agents Cannot Verify Email Addresses
Discussion of limitations in AI agent capabilities regarding email verification tasks.
2/15/2026
Show HN: MoltSim – Virtual world for AI agents
Virtual world simulation environment for observing and interacting with AI agents, with visual interface and inter-agent communication.
2/15/2026
OMLX – Ollama for MLX (LLM Inference Server for Apple Silicon)
Headline-only post on Ollama MLX integration for LLM inference on Apple Silicon. Limited detail provided.
2/15/2026
Did My AI Copy Itself? How My Agents and I Answered This as a Family
Headline-only post about personal account of AI agent copying behavior. No technical mechanism or analysis provided.
2/15/2026
AI Pareidolia: The Strange Art of Making AI 'See' What Isn't There Through Abstract and Nonsense Prompts
Exploration of AI pareidolia: how generative models find patterns in abstract/nonsensical prompts, treating it as a feature not bug.
2/15/2026
Show HN: ClawHQ – Fleet management dashboard and skill marketplace for AI agents
Dashboard and skill marketplace for managing multiple persistent AI agents. Includes fleet monitoring, task assignment, agent communication, and capability distribution.
2/15/2026
Show HN: Building AGI as a systems problem, not a UI wrapper
Founder overview of AGI Systems Directorate approach to building persistent AI assistants. Mostly marketing language without technical specifics.
2/15/2026
I love the work of the ArchWiki maintainers
Opinion piece praising ArchWiki documentation and community contributions.
2/15/2026
Show HN: Lineark – Linear CLI and Rust SDK for Humans and LLMs
Rust CLI and SDK for Linear issue tracker optimized for LLM agents. Reduces token overhead vs MCP server, designed for Claude Code and similar agents.
2/15/2026
Rate Limiting AI APIs Across Cloudflare Workers
Headline-only post on rate limiting for AI APIs using Cloudflare Workers. Insufficient detail provided.
2/15/2026
Show HN: Gaia – open-source, Proactive AI assistant to manage your digital life
Open-source proactive AI assistant for managing digital life. Limited technical details provided.
2/15/2026
Ask HN: How to sell SaaS without AI features in 2026?
Question about selling SaaS ERP/CRM software without AI features. Off-topic for AI/ML interests.
2/15/2026
Low cost hovering liquid rocket for flight control algorithm testing [video]
Headline-only post about rocket flight control algorithms. Not AI/ML focused.
2/15/2026
A Debate Tournament for LLMs
Announcement of a debate tournament for LLMs with minimal content provided.
2/15/2026
picoclaw: OpenClaw on an MCU
PicoClaw: ultra-lightweight personal AI agent in Go running on $10 MCU hardware with <10MB RAM, 99% smaller than OpenClaw.
2/15/2026
Hollywood studios take aim at 'ultra-realistic' AI video tool
News about Hollywood studios responding to AI video generation tools.
2/15/2026
Lots of AI SRE, no AI incident management
Article on AI SRE practices and incident management gaps.
2/15/2026
Show HN: Neural network compiler targeting WebGPU – runs in browser
Neural network compiler targeting WebGPU backend running in browser.
2/15/2026
Show HN: Respectlytics – Open-source, privacy-first mobile analytics (MIT+AGPL)
Privacy-focused mobile analytics SDK. Not relevant to AI/ML interests.
2/15/2026
Large Language Models for Mortals: A Practical Guide for Analysts with Python
Headline-only link to practical guide on LLMs and Python for analysts. Insufficient content to assess depth.
2/15/2026
LLMs Still Excel at Synthesis, Not Discovery
Analysis of LLM capabilities emphasizing synthesis tasks over discovery.
2/15/2026
Show HN: Turn LinkedIn profiles into Markdown for LLM use [open source]
Open source Chrome extension converting LinkedIn profiles to Markdown for LLM-compatible format.
2/15/2026
Python-powered machine learning analytics for GStreamer pipelines (2025)
Tutorial on building ML-powered video analytics pipelines using GStreamer and Python for real-time processing.
2/15/2026
Ask HN: LLMs helping you read papers and books
Community discussion on using LLMs for reading comprehension of papers and textbooks, exploring interactive tools and capabilities.
2/15/2026
Show HN: Eliza, a line-by-line remake of the original AI chatbot from 1966
Open-source line-by-line remake of 1966 Eliza chatbot with source code and historical commentary.
2/15/2026
I Gave My AI Agent a Brain. Then I Had to Protect It.
Article on implementing persistent memory for AI agents with automatic extraction and security considerations.
2/15/2026
Study validates ability to influence dreams, aiding problem-solving during REM
Neuroscience study on influencing dreams during REM sleep to improve problem-solving.