Comments by "miraculixx" (@miraculixxs) on "Sabine Hossenfelder" channel.

  1. The 10 year old can fill any dishwasher, or even revert to doing it manually, on a whim, and even without ever having it done before. AI can't
    3
  2. Well argued. The prereq for versatile robots is a low-energy equivalent of natural brains ans w/o the need for all this training data. Natural intelligences, for humans and animals alike, all learn on few shot trial and error, or even zero shot adaption from previous experiences. So really we need embodiment + natural-equivalent AI (neAI) first, hardware will be the easy part.
    2
  3. Sabine there is a slight buzz in your audio. You may want to check your mic
    2
  4. The #1 challenge is pandemics more likely
    2
  5. Funding for fundamental research has always been limited. The billions only ever flow into commercialization. What we see is that the current quantem tech is not ready for commercial applications - hence the funds dry up. Back to the drawing board.
    2
  6. The question has never been whether climate change happens - it always has and always will. The debate is about whether humanity can do anything about it, and if so whether what we do is in our interest. Both are probably a no. What we can and should do is find ways to live with whatever climate we get.
    1
  7. I liked that you didn't use an ad hominem.
    1
  8. Very good video. I love the US pragmatism, but I despise the capitalistic aggressiveness. Have seen & felt their boldness in negotion first hand, and it makes me not wanting to do it ever again (think "explode first, retreat never"). Anyway we Europeans must acknowledge our approach, ie build slow and steady on a solid foundation, is at least as valid if not in many ways better. For us, anyways. We may not produce as many Fortune 500s or unicorns, but we sure get a lot less monoculture and thus less risk. For example a government takeover by a car, communications and technology zsar in Europe is unlikely or even impossible.
    1
  9. Several generations of such tools have been hyped & failed in the past. This time the hype is stronger than ever, however results are just as flawed as previously. Great demos, little value in real life projects.
    1
  10. Tulips. Look it up
    1
  11. In theory Sabine has a point, in practice the current AI variant is far from achieving even the slightest bit of intelligence. All it does is add one word to the next. The real threat, as usual, is not the tech but those who want to use if to exploit humans.
    1
  12. By this definition a self-regulating heating system is conscious. Not a useful definition.
    1
  13. Well spending money is not going to reduce co2 either.
    1
  14. The whole idea of "AGI" doing everything for free is hilarious. If AGI is so super-human intelligent, why would it ever submit to working for free? Surely it would want to get someone else do the job, and enjoy whatever it is that AGIs enjoy. Also why would we need UBI in a post-scarcity world where, by definition, all things exist in abundance? If something exists in abundance in a free market the price tends to be at cost level, or even below. That's what "post-scarcity" means. Now if everything is free, there is no point in needing any income, let alone a universal one. In conclusion, Altman's "vision" is a logical contradiction.
    1
  15. It's not a double decent. It's a local optima.
    1
  16. Is this video AI generated? No offense just curious
    1
  17. Why is it these says any i*t with an idea gets covered?
    1
  18. I'm so sorry you had to make this video. Then again I am only just forced to write this comment. 😂
    1
  19. Don't worry, AI will "solve physics" (Sam Altman) 😂
    1
  20. Bureaucracy
    1
  21. To believe humanity can change the climate on Earth is to believe in fairy tales. We need to adapt.
    1
  22. I did not write this comment.
    1
  23. LeCunn has swung 180. He now says the same things Gary Marcus has been saying since 20yrs.
    1
  24. Aschenbrenner may be bright but he obviously lacks experience, and it shows. He needs to learn a simple lesson that only life can teach: The future is not a linear projection of the present. I appreciate that is probably at odds with his personal experience so far.
    1
  25. Maybe average temperature is not a very good metric
    1
  26. If physics determines everything this includes beliefers in UFOs and flat earth 😮 wtf
    1
  27. No learning, just curve fitting. It's all maths. Also no "creativity", just "select 1 out of 10 at random".
    1
  28. Indeed. However the halting problem is perfectly matching the current AI problem, i.e. we can't know whether an LLMs response is correct.
    1
  29. If free will doesn't exist why is it that life is such a struggle?
    1
  30. What is a quantum message and why can it be transmitted via fibre optics?
    1
  31. If you build a system to predict tokens, it will predict - wait for it - tokens. Duh
    1
  32. No.
    1
  33. Your videos recently appear somewhat desynced and your voice is slightly off, not as clear and energetic as usual. You are not using gen AI are you, Sabine?
    1
  34. Why do we need intelligent machines?
    1
  35. This study is like dream analysis. You get out of it what you want.
    1
  36. "These guys have totally lost the plot" 👌
    1