randomwalker

6 ms·

1. 1.▲
   AI agents can't yet do open-ended AI research(normaltech.ai)

   3 points by randomwalker 1mo ago|0 comments
2. 2.▲
   Can AI agents conduct open-ended AI research?(arxiv.org)

   5 points by randomwalker 2mo ago|0 comments
3. 3.▲
   What will be left for us to work on?(normaltech.ai)

   180 points by randomwalker 2mo ago|264 comments
4. 4.▲
   Up the Stack: How AI's Escape from the Commodity Trap Risks Enterprise Lock-In(normaltech.ai)

   3 points by randomwalker 2mo ago|0 comments
5. 5.▲
   Did Google's AI agents build an operating system for $916?(normaltech.ai)

   4 points by randomwalker 4mo ago|0 comments
6. 6.▲
   Open-world evaluations for measuring frontier AI capabilities [pdf](cruxevals.com)

   2 points by randomwalker 5mo ago|0 comments
7. 7.▲
   Towards a science of AI agent reliability(normaltech.ai)

   1 points by randomwalker 7mo ago|0 comments
8. 8.▲
   When AI Builds AI – Findings from a Workshop on Automation of AI R&D [pdf](cset.georgetown.edu)

   1 points by randomwalker 8mo ago|0 comments
9. 9.▲
   The Longitudinal Expert AI Panel: Understanding Expert Views on AI [pdf](static1.squarespace.com)

   1 points by randomwalker 10mo ago|0 comments
10. 10.▲
    Holistic Agent Leaderboard: The Missing Infrastructure for AI Agent Evaluation(arxiv.org)

    1 points by randomwalker 11mo ago|0 comments
11. 11.▲
    by randomwalker 1y ago

    Yes, we're aware! Fortunately our book is not a broad indictment of AI :) And none of our claims are premised on tasks people can do remaining out of reach for AI. More here: https://www.normaltech.ai/p/faq-about-t
12. 12.▲
    by randomwalker 1y ago

    Thanks! HN was part of the origin story of the book in question. In 2018 or 2019 I saw a comment here that said that most people don't appreciate the distinction between domains with low irreducible error that benefit from fancy models
13. 13.▲
    America's AI Action Plan [pdf](whitehouse.gov)

    11 points by randomwalker 1y ago|0 comments
14. 14.▲
    Could AI slow science? Confronting the production-progress paradox(aisnakeoil.com)

    2 points by randomwalker 1y ago|0 comments
15. 15.▲
    by randomwalker 1y ago

    Thanks for the comment! I agree — it's important to remain fluid. We've taken steps to make sure that predictively speaking, the normal technology worldview is empirically testable. Some of those empirical claims are in this paper
16. 16.▲
    by randomwalker 1y ago

    We do not assume a status quo or equilibrium, which will hopefully be clear upon reading the paper. That's not what normal technology means. Part II of the paper describes one vision of what a world with advanced AI might look like, an
17. 17.▲
    by randomwalker 1y ago

    I appreciate the concern, but we have a whole section on policy where we are very concrete about our recommendations, and we explicitly disavow any broadly anti-regulatory argument or agenda. The "drastic" policy interventions tha
18. 18.▲
    AI as Normal Technology(knightcolumbia.org)

    239 points by randomwalker 1y ago|92 comments
19. 19.▲
    Why an overreliance on AI-driven modelling is bad for science(nature.com)

    1 points by randomwalker 1y ago|0 comments
20. 20.▲
    Is AI progress slowing down?(aisnakeoil.com)

    5 points by randomwalker 2y ago|1 comments
21. 21.▲
    We Looked at 78 Election Deepfakes. Political Misinformation Isn't an AI Problem(knightcolumbia.org)

    5 points by randomwalker 2y ago|0 comments
22. 22.▲
    Inference Scaling FLaws: The Limits of LLM Resampling with Imperfect Verifiers(arxiv.org)

    3 points by randomwalker 2y ago|0 comments
23. 23.▲
    Is the UK's liver transplant matching algorithm biased against younger patients?(aisnakeoil.com)

    93 points by randomwalker 2y ago|62 comments
24. 24.▲
    Core-Bench: Computational Reproducibility Agent Benchmark(arxiv.org)

    1 points by randomwalker 2y ago|0 comments
25. 25.▲
    AI companies are pivoting from creating gods to building products(aisnakeoil.com)

    133 points by randomwalker 2y ago|195 comments
26. 26.▲
    AI Agents That Matter(aisnakeoil.com)

    35 points by randomwalker 2y ago|10 comments
27. 27.▲
    AI Agents That Matter(arxiv.org)

    4 points by randomwalker 2y ago|0 comments
28. 28.▲
    Scientists should use AI as a tool, not an oracle(aisnakeoil.com)

    124 points by randomwalker 2y ago|106 comments
29. 29.▲
    AI safety is not a model property(aisnakeoil.com)

    2 points by randomwalker 2y ago|0 comments
30. 30.▲
    AI safety is not a model property(aisnakeoil.com)

    3 points by randomwalker 3y ago|0 comments