4 ms·

Some of us have loooooong experience with AI (I've shipped product with AI features, and also have a 30Y-old AI degree) but also of marketing hype and outright

Some of us have loooooong experience with AI (I've shipped product with AI features, and also have a 30Y-old AI degree) but also of marketing hype and outright tech-based fraud (eg NFTs). There is some useful stuff going on, but outrageous self-serving wishful thinking by people with a track record of difficulty with reality makes it very hard to believe any of their promises and predictions.

arisAlexis 2y ago
[flagged]
- MultifokalHirn 2y ago
  So it baffles you that people trust their own experience and developed intuitions instead of just throwing them out because some big name people contradict them?
  - arisAlexis 2y ago
    Yes exactly. What I hear is physics teachers saying they know better than Bohr and Einstein.
    - foldr 2y ago
      Hinton (who's the only person on the list in your post who could really qualify as an expert here) has already made some glaringly wrong predictions about AI progress, such as the prediction that radiologists would be obsolete by now. No-one can predict the future. Experts have a slightly better track record than tea leaves – but only slightly. There are experts in $DISCIPLINE, but there are no experts on "the future of $DISCIPLINE".
      - arisAlexis 2y ago
        Of course AI is better than radiologists today but it will take some time to replace them. I give up though just wanted to see reactions.
        DamonHD 2y ago
        "Of course AI is better"? In all cases for all types of analysis and treatment? That is again trivially falsifiable. To be clear, I'm researching physically next to the AI people on campus and some of them are doing very smart stuff (including in medicine) but I don't think that any of the researchers or profs would repeat the casual claim that you just made.
      - arisAlexis 2y ago
        [flagged]
        foldr 2y ago
        [flagged]
        arisAlexis 2y ago
        [flagged]
        AnimalMuppet 2y ago
        If you're not going to listen to the answers, why did you ask the question?
        dang 2y ago
        Could you please stop posting unsubstantive comments and flamebait? You've unfortunately been doing it repeatedly, in this thread and others. It's not what this site is for, and destroys what it is for. If you wouldn't mind reviewing https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html and taking the intended spirit of the site more to heart, we'd be grateful.
    - DamonHD 2y ago
      Which actual teachers, when, with what words? There are idiotic people in all professions, and many of us make idiotic statements from time to time, but that in no way supports your main argument anyway.
    - appplication 2y ago
      Bohr and Einstein would be aghast at the idea of someone giving them credibility for who they are rather than questioning the merits of their ideas. It should perhaps give you pause that you’re implicitly conflating a group of arrogant capitalists with historical greats of science. Altman or Musk are not building or researching AI, they are marketing and selling it, and they will continue to do so in spite of the current or future state or capabilities. I would be careful not to confuse the loudest in the room with the smartest.
    - red-iron-pine 2y ago
      Bohr and Einstein did the math and had the proofs. and the math held up. Copilot has yet to give me a consistently correct set of "exploited in the wild" list of answers when I give it a bunch of CVEs. ChatGPT gave me last year's tax data for municipalities in Canada.
- DamonHD 2y ago
  Gates has plently of problems with reality (eg he spent years trying to deny / suppress the Internet as we now know it, along with many extremely questionable business practices and an utter disregard for user privacy and security IMHO), and a Nobel is very evidently no guarantee that all a winner's thoughts are correct. They may be right about aspects of AI or not. It baffles me that some people seem to succumb to the "argument from authority" fallacy.
- rsynnott 2y ago
  > Hinton is a Nobel prize winner. Oh, right, well, then, that makes him an expert on everything, then. Excuse me, I've just got to take a very large dose of Vitamin C; Linus Pauling assures me that it prevents/treats virtually everything. > Gates has no problems with reality either. I mean, Gates-era Microsoft was absolutely prone to flights of fancy. Like, for every hundred things that the industry seizes on as "this will change everything", maybe five ultimately become a thing at all, and maybe one actually does change everything.
  - daveguy 2y ago
    To be fair, he is a Nobel Prize winner for his AI research. But that doesn't mean anything since we are all speculating about the proximity of something that clearly doesn't exist -- AGI.

arisAlexis 2y ago

[flagged]

MultifokalHirn 2y ago
So it baffles you that people trust their own experience and developed intuitions instead of just throwing them out because some big name people contradict them?
- arisAlexis 2y ago
  Yes exactly. What I hear is physics teachers saying they know better than Bohr and Einstein.
  - foldr 2y ago
    Hinton (who's the only person on the list in your post who could really qualify as an expert here) has already made some glaringly wrong predictions about AI progress, such as the prediction that radiologists would be obsolete by now. No-one can predict the future. Experts have a slightly better track record than tea leaves – but only slightly. There are experts in $DISCIPLINE, but there are no experts on "the future of $DISCIPLINE".
    - arisAlexis 2y ago
      Of course AI is better than radiologists today but it will take some time to replace them. I give up though just wanted to see reactions.
      - DamonHD 2y ago
        "Of course AI is better"? In all cases for all types of analysis and treatment? That is again trivially falsifiable. To be clear, I'm researching physically next to the AI people on campus and some of them are doing very smart stuff (including in medicine) but I don't think that any of the researchers or profs would repeat the casual claim that you just made.
    - arisAlexis 2y ago
      [flagged]
      - foldr 2y ago
        [flagged]
        arisAlexis 2y ago
        [flagged]
        AnimalMuppet 2y ago
        If you're not going to listen to the answers, why did you ask the question?
        dang 2y ago
        Could you please stop posting unsubstantive comments and flamebait? You've unfortunately been doing it repeatedly, in this thread and others. It's not what this site is for, and destroys what it is for. If you wouldn't mind reviewing https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html and taking the intended spirit of the site more to heart, we'd be grateful.
  - DamonHD 2y ago
    Which actual teachers, when, with what words? There are idiotic people in all professions, and many of us make idiotic statements from time to time, but that in no way supports your main argument anyway.
  - appplication 2y ago
    Bohr and Einstein would be aghast at the idea of someone giving them credibility for who they are rather than questioning the merits of their ideas. It should perhaps give you pause that you’re implicitly conflating a group of arrogant capitalists with historical greats of science. Altman or Musk are not building or researching AI, they are marketing and selling it, and they will continue to do so in spite of the current or future state or capabilities. I would be careful not to confuse the loudest in the room with the smartest.
  - red-iron-pine 2y ago
    Bohr and Einstein did the math and had the proofs. and the math held up. Copilot has yet to give me a consistently correct set of "exploited in the wild" list of answers when I give it a bunch of CVEs. ChatGPT gave me last year's tax data for municipalities in Canada.
DamonHD 2y ago
Gates has plently of problems with reality (eg he spent years trying to deny / suppress the Internet as we now know it, along with many extremely questionable business practices and an utter disregard for user privacy and security IMHO), and a Nobel is very evidently no guarantee that all a winner's thoughts are correct. They may be right about aspects of AI or not. It baffles me that some people seem to succumb to the "argument from authority" fallacy.
rsynnott 2y ago
> Hinton is a Nobel prize winner. Oh, right, well, then, that makes him an expert on everything, then. Excuse me, I've just got to take a very large dose of Vitamin C; Linus Pauling assures me that it prevents/treats virtually everything. > Gates has no problems with reality either. I mean, Gates-era Microsoft was absolutely prone to flights of fancy. Like, for every hundred things that the industry seizes on as "this will change everything", maybe five ultimately become a thing at all, and maybe one actually does change everything.
- daveguy 2y ago
  To be fair, he is a Nobel Prize winner for his AI research. But that doesn't mean anything since we are all speculating about the proximity of something that clearly doesn't exist -- AGI.