Lobster Roll

All LHN/.

AI / Machine Learning

RSS
Loom vs. Linear: A tale of two AI-cities (jamespember.substack.com)
Family of Tumbler Ridge shooting victim suing OpenAI (cbc.ca)
Show HN: Open-source, model-agnostic alternative to Claude Code Review (github.com)
LLM Integrity During Inference in Llama.cpp (bednarskiwsieci.pl)
The Requirements Layer Your AI System Is Missing (kiln.tech)
Convert Any API Documentation into a CLI for AI Agents (github.com)
Anthropic sues Pentagon claiming supply chain risk label could cost billions (ft.com)
Porting MacPaint to Swift with Claude Code (weirdvibes.net)
Ask HN: Read‑only LLM tool for email triage and knowledge extraction?
I’m looking for an LLM‑powered tool to help organize and query my email, but with a strict boundary: it must not write, send, or modify anything.<p>Standard filters miss the nuance of &quot;middle-ground&quot; emails, and existing AI email clients demand full inbox control or try to automate too muc...
Ask HN: Any informed guesses on the actual size/architecture of GPT-5.4 etc.?
Does anyone have decent intuitions or hard clues on how big models like GPT-5.4, Gemini 3.1, and Opus 4.6 actually are, and how they compare to the best open models like GLM-5?<p>Are they all roughly in the same range now (for example around 1T params, maybe MoE), or are the closed models still much...
AI-powered 10-minute habit that taught my kid to read (and made me a better dad) (talperry.com)
Show HN: I built an AI-powered technical interview prep tool (crackr.dev)
LeetCode wasn&#x27;t actually preparing me or anyone else for the pressure of a real technical interview.<p>Most platforms just give you a text box and a compiler. I wanted something that actually simulates the back-and-forth of an interviewer pushing you on time complexity, asking follow-up questio...
Nvidia Is Planning to Launch an Open-Source AI Agent Platform (wired.com)
Claude Code Starter CLI (github.com)
No, it doesn't cost Anthropic $5k per Claude Code user (martinalderson.com)
The emerging role of SRAM-centric chips in AI inference (gimletlabs.ai)
Agentis – An AI-native programming language where the LLM is the stdlib (github.com)
Show HN: Free AI resume tailor I built after a recent layoff (300+ users so far) (jobbi.app)
Hey HN,<p>I’m a Staff Software Engineer, and just recently, on February 24, 2026, I was laid off along with 200+ coworkers. To survive the brutal job market, I hacked together a tool to automate my own applications: jobbi.app.<p>It just crossed 300+ users organically, so I wanted to share it here.<p...
From Tool to Employee: What Claude Code's /Loop Means (aieatingsoftware.substack.com)
Reinforcement fine-tuning use cases (developers.openai.com)
Making Prompt Injection Harder Against AI Coding Agents (medium.com)
Catching malicious contributions in open source repos (datadoghq.com)
New AI Note Tool
Hulunote – an outliner note app with a built-in MCP client
Head to head: Claude Code (Opus 4.6 / 1M) vs. Cursor (Composer 1.5 / 200k) (medium.com)
Managing My Open Source Repos with Autonomous AI Agents (massadas.com)
Aatel: The Anti-AI Training Ethical License – What It Is and Why It Was Built (github.com)
Sorca – Voice-first AI therapy companion (sorca.life)
Context plane for AI agents (Rust, S3) (aboutphilippe.com)
Anthropic says Trump ban puts federal contractor partnerships 'in jeopardy' (fedscoop.com)
Moody humans should let AI handle bad public feedback first, study finds (theregister.com)