Comments by "Cestarian Inhabitant" (@cestarianinhabitant5898) on "The first casualties of AI" video.

  1. It's a mixed bag. I've tried to make chatgpt wirte some code for me, pointed it out it's exact mistakes, and it said sorry, you're right, then puked out code with the same mistakes exacerbated.
    20
  2. ​ @Preinstallable  This is something that will happen, it is inevitable. An ai that can code as well as a human programmer, will be created by human programmers, only unlike human programmers it's not gonna make mistakes, it's not gonna need to debug, the human programmers worked for years to make this ai, and this ai is gonna be able to make a new superior version of itself in a couple of minutes. But will that newly created superior version be able to create a new even more superior version? Probably not. It just doesn't work like that. Until you get into AI running on quantum computers that is, then it's probably gonna work exactly like that. Yikes. When we mix technology so complex we barely understand it (neural networks), with another technology so complex we don't understand it at all (quantum physics). That's where this singularity is gonna happen, 100%. That's the missing piece of this puzzle, and you know what, it's not gonna be long from now, in fact I'm sure someone is already working on this. It's just that the quantum computers we have aren't ready yet for this task... But they will be, in just a few years they will inevitably be ready for this. Right now, quantum computers are growing exponentially in power by the year. It's growing so fast it makes moore's law back in the days of CPU transistors being half the size next year they were the last a funny joke. This is happening soon. All the AI tech we're using right now is gonna be obsolete within 10 years. Maybe even within 5....
    3
  3. Same old story. This is how it has always been.
    2