Video: ChatGPT Boxing Knowledge Experiment

Discussion in 'Classic Boxing Forum' started by Rumsfeld, Feb 16, 2023.


  1. Rumsfeld

    Rumsfeld Moderator Staff Member

    49,541
    16,033
    Jul 19, 2004
    This was the first time I've "improvised" a video in more than 6 years, and it wasn't especially well thought out. I'd nearly scrapped it putting it together, but figured the results proved interesting.

    Apparently Riddick Bowe and Lennox Lewis had not one, but TWO fights in the professional ranks, in an era defining rivalry. Who knew?

    Thought some of you may get a kick out of this. All in good fun, although I think this does expose the limitations of the AI in its current form.

    Open the pod bay doors, HAL!! Open the pod bay doors!

    This content is protected
     
  2. Barrf

    Barrf Boxing Addict Full Member

    5,624
    8,599
    Sep 19, 2021
    ChatGPT is a polite, well-spoken, articulate bullshitter. It seems to just make **** up if it doesn't know the actual answer.

    I've played with it in the context of boxing as well. I asked it to describe Louis-Schmeling 2 to me. It got to the point where it was describing the action in round 4....

    It is better at writing fictional boxing stories. I asked it to write a story about Joe Louis and Mike Tyson fighting. It described an action packed event in which Louis KO'd Tyson after a short brutal fight. I also asked it to write a story about a fictional celebrity boxing match between Abraham Lincoln and Barack Obama. It was clever enough to mention Lincoln's height and reach advantage, and describe Lincoln as an outfighter with a good jab. As far as the outcome, it gave Obama fast hands and had him repeatedly get inside to land combos until eventually he KO'd Lincoln.
     
    Last edited: Feb 16, 2023
    Jel likes this.
  3. FrankinDallas

    FrankinDallas FRANKINAUSTIN

    30,079
    36,890
    Jul 24, 2004
    As the Robot where the film for Greb-Tunney I is.
     
  4. salsanchezfan

    salsanchezfan Obsessed with Boxing Full Member

    15,800
    11,424
    Aug 22, 2004
    Careful what you tell it. It's collecting everything you say.

    As for published works, OpenAI, who owns and developed it, hasn't paid any attention at all to copyrights either. They glean material from any source they can, and have never paid a cent for reuse of it.

    It's a bad thing, and I lament it's presence.