Comments by "" (@diadetediotedio6918) on "are we cooked w/ o3?" video.

  1.  @BlackTakGolD  It is not stupid, if they achieved something revolutionary they would have good reasons to use it for themselves first.
    7
  2. We had massive layoffs since always, lol. I remember since 2020 people were complaining about this (when I started noticing, before there also this).
    2
  3. Maybe it is the fear of being replaced, or maybe it is the overconfidence of people blindly saying AI will replace programmers that drives the madness of our minds. In any way, I would not worry until something happens for a fact instead of speculating online ad nauseam.
    2
  4. The argument is not [only] the cost, the cost is a [part] of it.
    2
  5.  @steve_jabz  I don't think you are being faithful, he never stated he could see the reasoning of o3, did he? Because that earlier test of OpenAI itself on o1 about its safety said it was not able to confidently assert the thoughts of the model because they are private, if the OpenAI researchers cannot see it entirely (and I think it is humanly impossible to analyze 8B tokens to check every single possible word said and atest if it is or not bruteforced) what makes you think he could? Also, what I mean with bruteforcing is something much more particular: the model generates answers until one of them matches the provided examples patterns, this is a pretty reasonable way of a good pattern matcher (which AI's are) to identify if a given response can with high probability satisfy a given question.
    2
  6. He ranted about other things, bot.
    1
  7.  @Sa-jv7cl  Because I do this for almost 10 years, and I have used AI for work for 3, it is not that good.
    1
  8. I never once saw he changing his points, lol. What points he changed with the versions?
    1
  9.  @JArielALamus  It was never to "understand how our brain works", it was always to solve a problem.
    1
  10. I never saw a software engineer being fired "because AI", this don't even make sense.
    1
  11. ​ @umapessoa6051  No, it literally can not.
    1
  12.  @steve_jabz  It is literally written "TUNED" in caps in the graph, lol. I think it is also very bold to say a dataset don't leak any information about the private thing, this appears to be literally impossible to measure.
    1
  13.  @steve_jabz  You mean, this? lol [[[OpenAI shared they trained the o3 we tested on 75% of the Public Training set. They have not shared more details. We have not yet tested the ARC-untrained model to understand how much of the performance is due to ARC-AGI data.]]] And you only answered half of my comment.
    1
  14.  @steve_jabz  He did not proved that it is not brute force tho, how would he do this?
    1
  15. What about you betting money on your beliefs?
    1
  16.  @natsukisubaru2683  The point for those people is merely to prove a point, they lack the basis of any kind of deeper thinking than "huh, humans are not special, machines will replace us all", it is like cattle, but cattle at least can't talk.
    1