AI/ML links

• AI code is legacy code from day one167
• As an Experienced LLM User, I Actually Don’t Use Generative LLMs Often167
• I’d rather read the prompt167
• Time saved by AI offset by new work created, study suggests167
• A cheat sheet for why using ChatGPT is not bad for the environment.167
• The AI native developer landscape162
• Things you never dared to ask about LLMs162
• An Opinionated Guide on Which AI Model to Use in 2025162
• Making Sense of the AI Developer Tools Ecosystem162
• Gemini 2.5 Pro vs. Claude 3.7 Sonnet: Coding Comparison162
• Welcome to The State of the AI Native Dev Ecosystem survey!162
• MCP - It’s Hot, But Will It Win?162
• Playwright MCP* The case against conversational interfaces162163
• AI masters Minecraft: DeepMind program finds diamonds without being taught163
• 12 Graphs That Explain the State of AI in 2025 Stanford’s AI Index tracks performance, investment, public opinion, and moreEliza Strickland21 hours ago8 min read163
• AI crawlers cause Wikimedia Commons bandwidth demands to surge 50%163
• Welcome to the 🤗 AI Agents Course163
• The NLP Course is becoming the LLM Course!163
• Shut it all down? Microsoft research suggests AI usage is making us feel dumber – but you don’t need to panic yet156
• AI agents can be easily tricked into doing stupid things, study says156
• AI is Stifling Tech Adoption156
• I’m glad AI didn’t exist when I learned to code156
• Scientists took years to solve a problem that AI cracked in two days157
• Claude 3.7 Sonnet and Claude Code157
• Get coding help from Gemini Code Assist — now for free157
• OpenAI Researchers Find That Even the Best AI Is “Unable To Solve the Majority” of Coding Problems157
• Damning new AI study shows that chatbots make errors summarizing the news over 50% of the time — and this is the worst offender157
• People are using Super Mario to benchmark AI now158
• A New AI Worry: Many Young Coders No Longer Know How Their Code Really Works158
• Save the AI158
• Comparing of AI tools158
• Hallucinations in code are the least dangerous form of LLM mistakes158
• Here’s how I use LLMs to help me write code159
• State of Webdev AI159
• How much energy will AI really consume? The good, the bad and the unknown159
• How Much Are LLMs Actually Boosting Real-World Programmer Productivity?159
• How I Code With LLMs These Days159
• LLMs Don’t Know What They Don’t Know—And That’s a Problem159
• Embeddings (and how to create them)160
• ‘Open’ AI model licenses often carry concerning restrictions160
• Introducing AI Assistance in Chrome DevTools160
• Google for Startups Accelerator: AI for Energy applications are open160
• AI is reshaping UI — have you noticed the biggest change yet?160
• AI is the future of accessibility161
• Adobe Analytics: Traffic to U.S. retail websites from Generative AI sources jumps 1,200 percent161
• Open Source devs say AI crawlers dominate traffic, forcing blocks on entire countries161
• Gemini 2.5 is here, and it is an absolute beast161
• Gemini 2.5: Our most intelligent AI model161
• Gemini 2.5 Pro is here with bigger numbers and great vibes161
• Introducing 4o Image Generation161
• OpenAI brings native image generation to ChatGPT161
• Open source devs say AI crawlers dominate traffic, forcing blocks on entire countries161
• AI search is starting to kill Google’s ‘ten blue links’161
• Embeddings (and how to create them)160
• ‘Open’ AI model licenses often carry concerning restrictions160
• Introducing AI Assistance in Chrome DevTools160
• Google for Startups Accelerator: AI for Energy applications are open160
• AI is reshaping UI — have you noticed the biggest change yet?160
• If You Think Anyone in the AI Industry Has Any Idea What They’re Doing, It Appears That DeepSeek Just Accidentally Leaked Its Users’ Chats155
• Exciting updates to the Copilot (AI) Bounty Program: Enhancing security and incentivizing innovation155
• Growing Up AI-First155
• Customers don’t care about your AI feature155
• AI Makes Tech Debt More Expensive155
• AI is Creating a Generation of Illiterate Programmers153
• Thoughts On A Month With Devin153
• DeepSeek FAQ153
• Announcing the Web AI Acceleration Fund153
• Announcing the Web AI Acceleration Fund154
• A minimal PyTorch implementation for training your own small LLM from scratch.154
• Downloading DeepSeek could lead to jail time and a $1 million fine under new US bill154
• Google deletes policy against using AI for weapons or surveillance154
• Responsible AI: Our 2024 report and ongoing work154
• Introducing deep research154
• Google Ad pointing to tainted Homebrew install152
• Malicious extensions circumvent Google’s remote code ban152
• Cascading Spy Sheets: Exploiting the Complexity of Modern CSS for Email and Browser Fingerprinting152
• Malicious extensions circumvent Google’s remote code ban152
• OWASP’s New LLM Top 10 Shows Emerging AI Threats152
• Cloudflare location leak152
• Why AI Is Making Dev Skills More Valuable, Not Less146
• DeepMind’s Genie 2 can generate interactive worlds that look like video games146
• OpenAI releases AI video creator Sora but it won’t be coming to Europe yet146
• Devin is now generally available146
• Google DeepMind unveils Genie 2, an AI that Generates Playable 3D Worlds146
• The 70% problem: Hard truths about AI-assisted coding146
• Sora is here146
• Perplexity expands its publisher program146
• Certain names make ChatGPT grind to a halt, and we know why146
• Background image remover147
• How Nightshade Works147
• AI can’t read tables147
• Generating worlds147
• Summarizing with Transformers.js147
• Web AI Summit 2024: State of client side machine learning147
• Copilot Free147
• 15 Times to use AI, and 5 Not to147
• OpenAI’s API users get full access to the new o1 model147
• Using Transformers.js for AI in the Browser147
• A former Google CEO on when ‘we need to seriously think about unplugging’ AI148
• Most iPhone owners see little to no value in Apple Intelligence so far148
• Despite its impressive output, generative AI doesn’t have a coherent understanding of the world148
• Is AI Finally Starting to ‘Get’ Geography?148
• Putting AI to the (Accessibility) Test148
• AI will be dead in five years148
• The rise of the AI crawler149
• Cursor vs GitHub Copilot149
• ChatGPT search vs. Google: A deep dive analysis of 62 queries149
• Do Not Be Misled by ‘Build an App in 5 Minutes’: In-Depth Practice with Cursor149
• Ai Tools make me more productive, but i feel like i’m loosing my skills149
• Amazon Announces Nova, it’s New Family of Multimodal AI Models145
• Google’s new generative AI video model is now available145
• Spotify Wrapped 2024 adds an AI podcast powered by Google’s NotebookLM145
• Principles of Developer AI Product Development145
• Devs gaining little (if anything) from AI coding assistants145
• Argentina’s crime prediction paper128
• Argentina’s president wants to predict crime using AI algorithms128
• Olympics’ AI Security Stokes Backlash Over Mass Surveillance128
• At the Olympics, AI is watching you128
• There’s a Tool to Catch Students Cheating With ChatGPT. OpenAI Hasn’t Released It.128
• Google claims new AI training tech is 13 times faster and 10 times more power efficient — DeepMind’s new JEST optimizes training data for impressive gains124
• Explainer for the Prompt API
• A language model in your browser, accessible as a Web API?124
• Exclusive: This is Google AI, and it’s coming to the Pixel 9123
• Web content, the social contract and copyrights123
• Consent, LLM scrapers, and poisoning the well123
• Perplexity AI is susceptible to prompt injection123
• Declare your AIndependence: block AI bots, scrapers and crawlers with a single click
• New algorithm discovers language just by watching videos121
• A Right to Warn about Advanced Artificial Intelligence119
• What happened when 20 comedians got AI to write their routines121
• The last chatbot you will ever need121
• Chatty is your private AI that leverages WebGPU to run large language models121
• Entirely-in-browser, fully private LLM chatbot supporting Llama 3, Mistral and other open source models.121
• Built-in AI
• How does AI impact my job as a programmer?
• What We Learned from a Year of Building with LLMs (Part I)
• AI is already better at analyzing financial reports than humans
• InstructAvatar: Text-Guided Emotion and Motion Control for Avatar Generation
• Meredith Whittaker on AI
• Artificial intelligence technology behind ChatGPT was built in Iowa — with a lot of water
• Why you shouldn’t use AI to write your tests
• Jerky, 7-Fingered Scarlett Johansson Appears In Video To Express Full-Fledged Approval Of OpenAI
• No, Today’s AI Isn’t Sentient. Here’s How We Know
• Only about 5% of GenAI projects lead to significant monetization of new product offerings
• Detect AI-generated images & audio for your business - and yourself - with AI or Not
• What can LLMs never do?
• Sperm whale ‘alphabet’ discovered, thanks to machine learning
• How we developed Chrome’s first AI tools
• Prompt Library
• How To Harness Mouse Interaction Data For Practical Machine Learning Solutions
• AI that can predict how specific humans will look, act and feel could do the jobs of fashion models, focus group members and clinical trial participants
• A dataset with long, human-annotated English descriptions for 15k images
• The teens making friends with AI chatbots
• StreamMultiDiffusion: Real-Time Interactive Generation with Region-Based Semantic Control
• How does ChatGPT work? As explained by the ChatGPT team.
• Looking for AI use-cases
• Considerations for AI Opt-Out
• OpenAI introduces Batch API with up to 50% discount for asynchronous tasks112
• OpenAI batch API docs112
• 20 Popular Open Source AI Developer Tools112
• Reader converts any URL to an LLM-friendly input with a simple prefix
• What is AI
• Learning from you about AI
• The Anthropic Cookbook provides code and guides designed to help developers build with Claude, providing copy-able code snippets that you can easily integrate into your own projects.
• Historical analogies for large language models
• AI Ethical Framework: AI Usage Questions
• Shape of AI115
• Netflix’s AI-generated images in true crime documentary stir controversy113
• SwapAnything: Enabling Arbitrary Object Swapping in Personalized Visual Editing
• How Tech Giants Cut Corners to Harvest Data for A.I.
• Blind internet users struggle with error-prone AI aids
• Drive-by accessibility tweaks
• How screen readers read special characters: an update
• Don’t Be Fooled: Much “AI” is Just Outsourcing, Redux111
• Stability AI reportedly ran out of cash to pay its bills for rented cloudy GPUs111
• ‘It’s very easy to steal someone’s voice’: how AI is affecting video game actors
• A ‘Law Firm’ of AI Generated Lawyers Is Sending Fake Threats as an SEO Scam
• AI is coming for the OnlyFans chat industry
• Anthropic researchers wear down AI ethics with repeated questions
• Google Books Is Indexing AI-Generated Garbage
• Inside the shadowy global battle to tame the world’s most dangerous technology
• The lifecycle of a code AI completion
• The State of Generative AI, 2024111
• aider is AI pair programming in your terminal
• Diving Deeper into AI Package Hallucinations
• EvoEval: Evolving Coding Benchmarks via LLM
• Generative AI for Beginners (Version 2) - A Course
• Deepfake Generation and Detection: A Benchmark and Survey
• Has your open source code be indexed by AI
• ScreenAI: A visual language model for UI and visually-situated language understanding
• Under The Hood: How OpenAI’s Sora Model Works
• 8 Google engineers invented modern AI
• Discover The Best AI Websites & Tools
• AI infrastructure explained111
• Am I in The Stack?111
• How GitHub Copilot works111
• How GitHub engineers use Copilot111
• Best practices on using Copilot in your IDE:111
• Developers with AI assistants need to follow the pair programming model.111
• How AI is unfairly targeting and discriminating against Black people108
• ASCII art elicits harmful responses from 5 major AI chatbots108
• Jailbreaking LLMs with ascii art
• Learn how AI works from a real LLM implemented entirely in Excel
• Clarity-Upscaler: Reimagined image upscaling for everyone
• Leonardo.Ai performantly generates 4.5 million images daily with Next.js and Vercel
• This dataset contains more than 650,000 English books (~ 61 billion words) presumed to be in the public domain in the US which were digitised by the Internet Archive and catalogued as part of the Open Library project. The dataset was compiled by Sebastian Majstorovic.
• “AI will cure cancer” misunderstands both AI and medicine
• Gemini is able to generate Selenium code to replicate tasks from a recorded screen capture.106
• AI Prompt Engineering Is Dead Long live AI prompt engineering106
• Top AI researchers say OpenAI, Meta and more hinder independent evaluations
• AI and democracy
• “AI” and accessible front-end components: is the nuance generatable?
• How Allowing Copyright On AI-Generated Works Could Destroy Creative Industries
• Dispelling the Generative AI fear: how Cloudflare secures inboxes against AI-enhanced phishing
• ‘He checks in on me more than my friends and family’: can AI therapists do better than the real thing?
• Google brings Stack Overflow’s knowledge base to Gemini for Google Cloud
• Happy New Year: GPT in 500 lines of SQL
• FlowMDM: Seamless Human Motion Composition with Blended Positional Encodings
• Introducing Phind-70B – closing the code quality gap with GPT-4 Turbo while running 4x faster
• Things I Don’t Know About AI
• Google’s New AI Can Generate Entire 2D Platformer Games105
• Jensen Huang says kids shouldn’t learn to code — they should leave it up to AI105
• I worry our Copilot is leaving some passengers behind105
• Question Answering Over 1 Hour Video.
• The digital equivalent of wearing a fake Chanel bag
• OpenAI’s next AI product could be after your job
• Extreme Video Compression with Pre-trained Diffusion Models
• GPTScript is a new scripting language to automate your interaction with a Large Language Model (LLM), namely OpenAI.
• Google Launches €25 Million AI Opportunity Initiative for Skills Training Across Europe
• Stripping the web of its humanity
• The LLM search engine
• What are Vectors, and how they are used in Machine learning
• BRIA Background Removal v1.4 Model Card
• When Words Cannot Describe: Designing For AI Beyond Conversational Interfaces
• The LLM Misinformation Problem I Was Not Expecting
• How to Learn Unfamiliar Software Tools with ChatGPT
• Ask HN: What have you built with LLMs?
• It’s easier to forgive a human than a robot
• AI Design Patterns
• How hard is it to cheat in technical interviews with ChatGPT? We ran an experiment.102
• You Are Not Late to Artificial Intelligence
• Welcome to the GPT Newspaper project, an innovative autonomous agent designed to create personalized newspapers tailored to user preferences.
• CapHuman: Capture Your Moments in Parallel Universes
• Using AI for 3D rendering—a practical guide for designers
• Why the AI Boom is a Windfall for Tiny Anguilla The Caribbean island is reaping millions from .ai website registrations
• AI today and trends for an AI future
• Prompts Are Tiny Programs
• Google shows off Lumiere, a space-time diffusion model for realistic AI videos
• Google’s latest AI video generator can render cute animals in implausible situations
• Google’s New AI-Powered Browser Could Mark the End of the Human Internet
• AI art is the new stock image
• Text-to-Video: The Task, Challenges and the Current State
• The Intelligence Illusion: stepping into a pile of ‘AI’
• Lies, damned lies, and benchmarks
• Generating code was never the hard part
• Prompt engineering with spreadsheets
• Do Users Write More Insecure Code with AI Assistants?101
• Simon Willison interview: AI software still needs the human touch101
• Telling GPT you are scared gives better results
• A list of known AI agents on the internet
• How Microsoft found a potential new battery material using AI
• A developer’s second brain: Reducing complexity through partnership with AI
• Upcoming Accessibility Events (2024)
• llamafile lets you distribute and run LLMs with a single file
• Dataset Card for Steamboat Willy frames
• Onboard.ai
• A guide to designing and shipping AI developer tools
• Adversarial Attacks on LLMs
• AI ‘Nudify’ Apps That Undress Women in Photos Soaring in Use - Bloomberg
• Clippy returned (as an unnecessary “AI”)
• David Attenborough narrates your life.
• Demystifying LLMs: How they can do things they weren't trained to do - The GitHub Blog
• Google’s DeepMind ‘Robot Constitution’ will make sure its AI droids won’t kill you - The VergeThe data gathering system AutoRT applies safety guardrails inspired by Isaac Asimov’s Three Laws of Robotics.
• How to choose a security tool for your AI-generated code
• Images altered to trick machine vision can influence humans too - Google DeepMind New research shows that even subtle changes to digital images, designed to confuse computer vision systems, can also affect human perception
• In Defense of AI Hallucinations
• Instant Voice Cloning
• Jan is an open-source ChatGPT alternative that runs 100% offline on your computer.
• Large Language Model Course
• NCSS Guidelines for secure AI system development
• The AI trust crisis
• The architecture of today’s LLM applications
• The I in LLM stands for intelligence daniel.haxx.se
• Web developers: remarkably untalented and careless? – Baldur Bjarnason
• WhiteRabbitNeo is a model series that can be used for offensive and defensive cybersecurity.
• State of AI Code Generation in 202397
• NSFW Filter An npm library that helps filter out inappropriate images using AI.
• Debunking the Myth of “Anonymous Data
• AI and the automation of work
• Stuff we figured out about AI in 2023
• Generative AI Tools for Developers: A Practical Guide
• AI and Lossy Bottlenecks
• 2023: The Year of AI
• Using Generative AI to Detect Cat Breeds97
• Restoring old photos using AI
• Apple’s ‘Ferret’ is a new open-source machine learning model
• Semantic Based Painting
• Animatable Gaussians - create avatars from videos93
• Train ChatGPT on your documentation93
• DiffEdit: Editing Images using Generative AI
• Coframe/coffee: Build and iterate on your UI 10x faster with AI
• Generative AI for Beginners
• ✨ 7 AI Libraries EVERY Dev Needs to Know (to be a wiz)🧙‍♂️ 🪄 - DEV Community
• Ask HN: Daily practices for building AI/ML skills? - Hacker News
• Dover - AI Applicant Sorting
• Prompt engineering - OpenAI API
• If AI is making the Turing test obsolete, what might be better? - Ars Technica
• AI-Powered Microdisplay Adapts to Users’ Eyesight - IEEE Spectrum
• 🗣️ Large Language Model Course
• Navigate the unexpected: using Copilot in Edge and DevTools - Microsoft Edge Blog
• Google debuts Imagen 2 with text and logo generation - TechCrunch
• Three types of AI-assisted programmers - Stack Overflow
• Bash One-Liners for LLMs
• Now add a walrus: Prompt engineering in DALL-E 3
• After 6 years Microsoft launches Seeing AI app for Android - gHacks Tech News
• Rethinking Compression: Reduced Order Modelling of Latent Features in Large Language Models
• krishnaik06/Roadmap-To-Learn-Generative-AI-In-2024
• Vector Databases: from Embeddings to Applications - DeepLearning.AI
• Why you should add friction to your onboarding
• Prompts are unsafe, and that means language models are not fit for purpose