Comments by "L.W. Paradis" (@l.w.paradis2108) on "Sam Altman POACHED By Microsoft, OpenAI Employees REVOLT Against Board" video.

  1. So far, AI processes immense quantities of data, far beyond what people, assisted with standard computers, could do in several hundred years. The analysis is purely statistical. That's why it has started making serious errors --- in mathematics, ironically enough. In the case of wrongly identifying prime numbers, which ChatGPT often does, it is probably because it is generalizing over millions of opinions about whether a given number is prime. 😂 You've read about this --- right?
    3
  2. ChatGPT has started making inexplicable errors in mathematics, including flubbing simple questions such as whether a given 6-digit number is prime. Not only that, but when you ask it to show its steps, it has stopped doing so.
    2
  3. Major law firms? They have their own proprietary templates. You knew which firm a letter was from, and lawyers were so hyper about style and firm identity that they debated whether to use "Yours truly" or "Sincerely" as a general closing, and who should get "Very best regards" instead. Don't get me started about pagination and mastheads. There is no way ChatGPT could be better at writing contracts than a major firm with a decades-long archive. A smaller one, sure, but not a major firm. The point of a contract is to write it to foreclose litigation. You need to know a lot of law to do that.
    2
  4.  @CoachAndrew77  But it doesn't know it doesn't know. It needs you to point it out.
    1
  5.  @CoachAndrew77  You know if you don't know the answer to a math problem . . . Right? LOL
    1
  6.  @CoachAndrew77  I doubt that. You know when you don't know the answer. That sort of reflexivity is characteristic of human consciousness. Don't underestimate it. It's the whole key to real intelligence.
    1
  7. @CoachAndrew77  Computers already can assist in identifying prime numbers. The largest known prime has over 82,500,000 digits, as of last month.
    1
  8. ​ @scratchpenny I will.
    1
  9.  @scratchpenny  Future generations of mankind? Or of Americans? Western Europeans? Five Eyes? Eastern Europeans? China, Japan, India, Vietnam, Brazil, Nigeria, Somalia? Were you told "no one" will learn cursive? Except that the other major languages of this hemisphere (Spanish, Portuguese, French) all use cursive. This sounds a lot like that.
    1
  10.  @CoachAndrew77  Wow, you've bought the propaganda, lol.
    1
  11.  @scratchpenny  "Arguing against?" I'm pointing out your narrow view. You may have a privileged perch from which to survey the impact of AI on the top three quintiles of Americans, classified by income, and of other Western(ized) speakers of English, and perhaps somewhat beyond that group. This doesn't give you any special perspective on what the rest of the planet will be doing, especially considering that you think AI's "virtues" will cause it to become pervasive and dominant. Well that's not how the world works. All of this rests on a different foundation, and that one is shaky.
    1
  12.  @scratchpenny  "It's going to have an outsized effect on how people perceive reality." --- Which people?
    1
  13.  @scratchpenny  "Future generations will be entirely dependent on it." Future generations where? In the BRICS?
    1