Hacker News
- Deep Learning, Deep Scandal https://garymarcus.substack.com/p/deep-learning-deep-scandal 21 comments
Linked pages
- Recent AI model progress feels mostly like bullshit — LessWrong https://www.lesswrong.com/posts/4mvphwx5pdsZLMmpY/recent-ai-model-progress-feels-mostly-like-bullshit 458 comments
- https://www.cs.mcgill.ca/~jpineau/ReproducibilityChecklist.pdf 29 comments
- Chatbots Are Cheating on Their Benchmark Tests - The Atlantic https://www.theatlantic.com/technology/archive/2025/03/chatbots-benchmark-tests/681929/ 6 comments
- https://old.reddit.com/r/LocalLLaMA/comments/1jt8yug/serious_issues_in_llama_4_training_i_have/ 3 comments
- Satya Nadella and the three stages of scientific truth https://garymarcus.substack.com/p/satya-nadella-and-the-three-stages 0 comments
- OpenAI says it'll release o3 after all, delays GPT-5 | TechCrunch https://techcrunch.com/2025/04/04/openai-says-itll-release-o3-after-all-delays-gpt-5/ 0 comments
Related searches:
Search whole site: site:garymarcus.substack.com
Search title: Deep Learning, Deep Scandal - by Gary Marcus - Marcus on AI
See how to search.