you are viewing a single comment's thread
view the rest of the comments
[–] 133 points 8 months ago* (9 children)

It's time to recalibrate my gradient on the big picture.

I guess this is AI-insider wit, but I'm so glad not to think or speak like these people do.

  • source
  • hideshow 9 child comments
  • [–] 55 points 8 months ago (2 children)

    It is, gradient descent is what you use to find optimal model parameters. 

    the algorithm takes a step, computes a gradient (whether any nearby options are better), then moves in that direction to improve the parameters, in a loop. 

  • source
  • parent
  • hideshow 2 child comments
  • [–] 5 points 8 months ago (1 child)

    Adding to the above, one of the challenges of GD is how to know whether the global optimum reported by the function isn’t just one of its many imposters (local optima). That’s the “big picture” he’s talking about. Working with Elon was a dead end.

  • source
  • parent
  • hideshow 1 child comment
  • [–] 5 points 8 months ago

    It seems obviously so. I don’t think I could hire someone who worked on Grok’s deep fake porn engine, ever. 

    Between the working for a nazi, child porn issue, and all, it’s a bad fucking look. 

  • source
  • parent
  • [–] 11 points 8 months ago (1 child)

    Someone from a brand new account posted a bunch of gibberish like that today about having the keys to the octo-dimension mother universe...

    And I immediately thought it was an Elon AI, because that's how they think humans actually talk

  • source
  • parent
  • hideshow 1 child comment
  • [–] 8 points 8 months ago*

    It’s language for people who advertise they know something about “AI,” but couldn’t implement it if their life depended on it.

    TBH I’ve never heard that one, but it sounds like they’re trying to use “gradient descent” in a sentence.

  • source
  • parent