🦞🌯 Lobster Roll

All LHN/.~Ars
RSS

Showing stories from 2025-01. View all

NewestOldestTop ScoredMost Discussed
Show HN: MyLog.food – AI-Powered Food Tracking Using Natural Language / Photos (mylog.food)
Reid Hoffman raises $25M to take on cancer with AI (qz.com)
Show HN: Couscous – AI-powered code reviews that keep your team in sync (github.com)
Couscous is a VS Code extension that uses AI to analyze your code quality against best practices and team conventions. It provides instant visual feedback, rewarding you with the satisfaction of delivering high-standard "code dishes" or... well, letting you know when your code needs a litt...
AI-Generated Code Is a Ticking Time Bomb (and It's Not the AI's Fault) (alexstandiford.com)
Simulating 500M years of evolution with a protein LLM (2024) (biorxiv.org)
Show HN: Aquila a small code embedding tool (github.com)
It's a small utility tool to embed code excerpts in your documentation.
OpenAI O3-Mini (openai.com)
OpenAI launches o3-mini, its latest 'reasoning' model (techcrunch.com)
Show HN: iPad app as your AI keyboard (flowy.live)
Hey folks, just dropping an early prototype of a project I’ve been working on.<p>Please watch the demo &amp; read the memo; it’ll tell you what v1 is about and where we’re headed.<p>I’m sorry that I couldn’t publish&#x2F;ready to install; the installation is janky, so please put your email down firs...
U.S. Investigates Whether DeepSeek Smuggled Nvidia AI GPUs via Singapore (tomshardware.com)
DeepSeek Is a Rorschach Test Revealing Each's AI Hopes and Fears (dbreunig.com)
2025 GitHub Copilot vulnerabilities – technical overview (apexhq.ai)
China created AI model DeepSeek and shocked the world (nature.com)
How much economic growth from AI should we expect, how soon? (inferencemagazine.substack.com)
I agree with OpenAI: You shouldn't use other peoples' work without permission (arstechnica.com)
Brazil's push for comprehensive AI regulation (restofworld.org)
How to have a career even when OpenAI's o3 drops (pradyuprasad.com)
I Use AI as an Intern (lucumr.pocoo.org)
Spy vs. AI: How Artificial Intelligence Will Remake Espionage (foreignaffairs.com)
Theoretical limitations of multi-layer Transformer (arxiv.org)
Copyright Office AI Report Says Creative Prompting Doesn't Constitute Authorship (ipwatchdog.com)
Transformer Crisis: Deepak Divan Sounds the Alarm (spectrum.ieee.org)
DeepSeek R1: Don't Put All Your Eggs in One LLM Basket (notdiamond.ai)
Big misconceptions of training costs for DeepSeek and OpenAI (old.reddit.com)
Ask HN: What are the best business AI-calls in Spanish?
I&#x27;ve tried them all (Bland, etc etc) but in spanish, they aren&#x27;t great!
Analyzing DeepSeek's System Prompt: Jailbreaking Generative AI (lab.wallarm.com)
I tested DeepSeek against domestic AI chatbots to see if Nvidia is screwed (sfgate.com)
Running NPU-Optimized DeepSeek R1 Models Locally on Copilot+ PCs (blogs.windows.com)
Colorado AI VIP: Dave Mayer, connector of people and community builder (coloradoai.news)
DeepSeek ripped up the AI playbook–and why everyone's going to follow it (technologyreview.com)