Comments by "clray123" (@clray123) on "ThePrimeTime" channel.

  1. The first time I have to agree with AI-driven censorship... I mean... C++ and kids... come on! Has science gone too far!?
    5
  2. I'd say they've been doing more of the left thing lately.
    4
  3.  @keyboard_g  The guy earns his money from blabber, not from code.
    4
  4.  @Alex-bx3st  Ah but it does. Stripping comments from existing code willy-nilly is the most obvious.
    3
  5. He is a "YouTube influencer". He does not code for a living.
    2
  6. I don't see a contradiction. They are not selling their product as a job application spam aid. In fact, I would not be surprised if they included it as an unallowed use case in their ToS. And it is reasonable that they want to find people who can solve problems with or without AI. Otherwise they would be hiring the AI instead (which may happen at some point, but we're just not there yet for the cutting edge research positions).
    2
  7. I believe provocation is part of their mission (pissed off people more happy to fund the military and support other such great projects).
    1
  8. I imagine your entire conversation is automatically flagged and some low-paid Google slave has to review it later for inappropriate content.
    1
  9. What do you expect from a language which teaches you about "destructors"?
    1
  10. COME to think of it, it does sound like a strange fetish...
    1
  11. Inherited from its creators...
    1
  12. Clearly it is the language of oppression. I suggest future AI to only answer in ebonics or Jamaican patois to promote diversity.
    1
  13. An endless series of mistakes, criticism, and apologies has been the usual "AI experience" for me. It feels a bit like hiring a stubborn idiot junior dev and having to supervise and teach him how to code... except this stubborn junior dev does not learn anything.
    1
  14. The point is that AI is actually usable now. Whereas most people's first contact was with the unusable AI. So it makes a lot of sense to whip the naysayers into giving it a second try. Because now it actually fucking works (for the most part, talking about Gemini 2.5 here).
    1
  15.  @BaneHydra  Neither Gemini 2.5 nor DeepSeek R1 included Mississipi (or any other invalid names). Both missed some states. As would you. If you want the list to be 100% accurate, do the same as you would do yourself, and let the model filter the actual list, rather than asking it to recite from memory.
    1
  16. @ I want to see you use Google Search to vectorize an edge detection algorithm written using opencv2 by porting it to pytorch and make it 3x faster in the process. You have 5 minutes. Good luck.
    1
  17. Good luck speaking to anyone @Google (unless you're their invited guest).
    1
  18. How is it a danger? Most people don't know assembler (any more), do you feel endangered by your lack of "assembler memory and skill"?
    1
  19.  @vinterskugge907  I'm pretty sure Russian hackers use C++. Think about it. (And think of the children, as usual.)
    1
  20. It is kids like you for whom we have juvenile detention.
    1
  21.  @xaviernogueira  The general theme with AI is that it solves problems it either created itself or which nobody has asked it to solve in the first place.
    1
  22. HAL9000, 1968: "I'm sorry Dave, I'm afraid I can't do that." Google, 2024: YES!
    1