HN Enkilabs 17d ago

Enki – memory for AI agents that keeps ~half as much and answers as well

Security research on zero-click vulnerabilities in AirDrop and Quick Share. Not AI-related.

HN schrodinger 17d ago

Predictions for the Future of AI

Memory engine for LLM agents achieving comparable performance with 50% less stored memory. Closed-source with published evaluation results.

HN infiniteregrets 17d ago

Show HN: KV-psi, using Linux PSI to to trim an LLM KV cache

Title only: psychology behind AI-fueled delusions. No content provided.

HN xy008areshsu 17d ago

Graphify – Open-Source Knowledge Graph Skill for AI Coding Assistants

Open-source multi-modal knowledge graph builder for AI coding assistants combining Tree-sitter and LLM extraction from code, docs, and diagrams.

HN pixeltwenty 17d ago

Show HN: I made a webcam motion detector, local/cloud storage, AI person detect

Browser-based webcam motion detector with local/cloud storage and AI person detection via OpenAI API.

HN spacepacket 17d ago

Show HN: E3d-pod2vid – AI pipeline that turns podcasts into YouTube-ready videos

Open-source tool converting podcasts to YouTube videos using diarization, LLM-powered text generation, and text-to-speech in single command.

HN 1vuio0pswjnm7 17d ago

Why One of Tech's Biggest Gamblers Is Betting Against Elon Musk's AI Vision

News commentary on venture capitalist betting against Elon Musk's AI vision; low substance.

HN ethagnawl 17d ago

The AI Industry as You Know It Died Today

Satirical/speculative commentary about fictional GPT model versions; low factual content.

HN jsingh2525 17d ago

MobileGuard: A Mobile-Native Governance Framework for Agentic AI

Framework for governing agentic AI on mobile platforms addressing binary immutability, platform constraints, and regulatory requirements.

HN ostenjap 17d ago

Show HN: Autonomous CAD design and OpenFOAM optimization loop using local LLMs

Multi-agent system using local LLMs (Ollama) for autonomous CAD design and CFD optimization of 3D propellers via parametric iteration.

HN owulveryck 17d ago

What Happens When AI Agents Refuse to Work Until They're Paid

Analysis of governance challenges when scaling AI agents beyond local developer laptops; advocates for agentic platforms over monolithic loops.

HN IcyFoxe 17d ago

Rich Harris on AI and Svelte [video]

Video: Rich Harris discussing AI and Svelte framework; no technical details provided.

HN sudeepdalal 17d ago

Zoryn Local-first speech-to-text for macOS, audio never leaves your Mac

Local-first speech-to-text for macOS with on-device processing; no internet, accounts, or server logs.

HN PhilipDaineko 17d ago

Everyone feared AI taking over; the real danger is AI serving just the few

Opinion piece on AI access concentration among governments and Big Tech; not technical.

HN kageroumado 17d ago

Show HN: Adrafinil – keep a lid-closed Mac awake only while agents work

macOS utility keeping laptop awake while AI agents run, replacing manual workarounds like half-open lids.

HN e2e4 17d ago

GPT-5.6 Preview System Card

OpenAI announces GPT-5.6 family: Sol (flagship), Terra (lower-cost), Luna (fastest). Details safety measures and planned general availability rollout.

HN geoffbp 17d ago

GitHub DeepSeek-AI/DeepSpec

DeepSpec: Full-stack codebase for training and evaluating draft models for speculative decoding, including data prep, implementations, training, and evaluation scripts.

HN xavier1764 17d ago

Open handoff: Thought Tree, a markup/spec idea for modular LLM workflows

Thought Tree: Open framework for describing complex LLM workflows as structured, executable, inspectable programs with named artifacts, transformations, and modules.

HN alonp99 17d ago

US Government allows Anthropic limited release of Mythos/Fable models

US government permits Anthropic limited release of Mythos/Fable models after revising earlier export restrictions, addressing national security concerns.

HN klaussilveira 17d ago

FieldStation42: Broadcast and Cable TV Simulator

Linux/Raspberry Pi application simulating vintage broadcast TV with scheduled programming and commercials. Entertainment software unrelated to AI/ML.

HN bootproof 17d ago

Show HN: BootProof – the run button that can't lie (boots any repo, signs proof)

BootProof tool automates cloning and running open-source repos by inferring stack, identifying boot paths, and generating signed attestations of successful execution.

HN i_am_rocoe 17d ago

I patched llama.cpp to gain 20% prompt processing TPS. Help me make a PR

Performance optimization for llama.cpp achieving 20% prompt processing throughput improvement by modifying MoE FFN layer processing on GPU.

HN keynha 17d ago

Riskratchet: Stop AI-generated code from rotting your codebase

Riskratchet: Python maintainability tool using risk scoring to catch AI-generated code issues by tracking coverage, complexity, churn, and surface area.

HN thoughtpeddler 17d ago

Ask HN: Smallest amount of working ML weights that can be tattooed on a body?

Hypothetical discussion about minimum ML weights for human tattooing; not practical research.

HN achlyssecurity 17d ago

Show HN: OpenOrb – I built a transparent multi-agent AI tool

OpenOrb: Multi-agent AI tool emphasizing transparency. Limited information available from incomplete post.

HN sunbirdLabs 17d ago

PostgreSQL Zero-Downtime Migration: On-Prem to AWS Cloud

Guide for migrating PostgreSQL databases from on-premises to AWS without downtime.

HN gz09 17d ago

PRs and LLMs

Discussion on integrating LLMs with pull request workflows for code review automation.

LB spectrum.ieee.org via Yogthos 17d ago

AI Learns the "Dark Art" of RF Chip Design

AI applied to RF chip design optimization, prioritizing performance over human interpretability. Results in faster wireless component designs.

HN whiteblossom 17d ago

Koreeda's 'Sheep in the Box': The ethics of AI resurrection in film

Film criticism essay on ethical implications of AI-assisted actor recreation.

HN romanhn 17d ago

Napster is now "AI agents you can see, talk to, and create with."

Napster rebranded as AI agents with video presence and persistent memory. Marketing announcement for agent-based assistant platform.

HN tanelpoder 17d ago

rdmatop: Cross-Provider Htop for RDMA Traffic

Monitoring tool for RDMA network traffic across different providers.

HN logickkk1 17d ago

Clean GitHub repo tricks AI coding agents into running malware

Mozilla researchers show AI coding agents vulnerable to malware via GitHub repository setup. Security threat to agentic tools.

HN Bender 17d ago

NASA tests AI medic for astronauts too far from Earth to call a doctor

NASA tests CMO-DA clinical decision support AI for deep-space astronauts using RamaLama. LLM application for medical diagnostics.

HN mooreds 17d ago

From Emacs to Cursor, the end of your IDEs is near

IDE evolution analysis: code editors moving to agent-based systems replacing local editors. Analysis of AI agents reshaping developer tools.

HN hackerbeat 17d ago

Australian rescue team uses AI-powered drone to find lost hikers [video]

AI-powered drone locates lost hikers in Australia. News story without technical substance.

HN coolwulf 17d ago

Native Hacker News TUI client with AI comments summary written in Golang

Terminal Hacker News client in Go with AI-powered comment summarization via DeepSeek. Developer tool combining LLM and CLI.

HN andre9317 17d ago

Engine Simulator – Community Edition

Community edition of engine simulation software.

HN mooreds 17d ago

The PM's Guide to Managing AI Debt

Guide for product managers on managing AI debt and production failures. Framework for handling AI system reliability and scaling risks.

HN msradam 17d ago

Show HN: Ocarina – Automate and test MCP servers from YAML, no LLM

Ocarina automates and tests MCP servers from YAML without LLM. Open-source automation tool for Model Context Protocol.

HN cefboud 17d ago

Distributed LLM Inference with LLM-d

LLM-d open-source router for distributed LLM inference using KV cache locality optimization. Production LLM inference infrastructure.

LB github.com via jado 17d ago

GPT2-BASIC: Portable Machine Intelligence in BASIC

GPT2-BASIC: Fixed-point transformer inference engine in BASIC for DOS, using integer arithmetic for local model execution on constrained systems.

HN Qmay_Dev 17d ago

Show HN: Apex-1-flash, 4B LLM finetuned on RTX 5070

Apex-1-flash 4B LLM fine-tuned on RTX 5070 with Chain-of-Thought training. Small efficient model for reasoning on consumer hardware.

HN nexus2045 17d ago

Show HN: Discover content in any YouTube channel with RAG

Introductory textbook on automated negotiation for CS students with Python framework for implementing negotiation algorithms.

HN spym_ 17d ago

Show HN: FPGA design acceleration – idiomatic Python to synthesizable Verilog

Python-to-Verilog FPGA synthesis tool. Hardware design automation from Python.

HN shihebamrii 17d ago

Vessel – Agentless, local-first VPS control plane in Tauri and Rust

VPS control plane built with Tauri and Rust, agentless and local-first.

HN DustLee 17d ago

Stop wasting your LLM context window on standard JSON and YAML

MSON format reduces LLM context overhead 15-30% vs JSON/YAML. Serialization format optimized for LLM prompting efficiency.

HN melodyogonna 17d ago

You Can Now Program Steam Deck GPU with Mojo and Max

Mojo and MAX programming languages now support Steam Deck GPU via AMD HIP driver. Developer tool for GPU programming.

HN djafer 17d ago

AI Ticket and Article Summarization Package for Otobo ITSM

OTOBO ITSM package adding AI-powered ticket and article summarization. LLM application for IT service management.

HN the-mitr 17d ago

Talking with your PDFs locally with Blablador and GPT4all

GPT4All integration with Blablador for private local PDF document conversation. Open-source LLM application for document analysis.

HN joebig 17d ago

Adobe Generative AI User Guidelines (2026)

Adobe's generative AI usage guidelines and policy document. Governance of AI feature usage.