"AI is so complex even the developers dont understand it" "looks inside" "if-statement and for-loops" cat stares down at the camera

all 29 comments

sorted by: hot top controversial new old
[–] 15 points 3 hours ago
[–] 36 points 4 hours ago (2 children)

That was old AI.

New AI uses a complex, multi-layer sequence of Magic 8-balls.

  • source
  • hideshow 2 child comments
  • [–] 17 points 4 hours ago*

    Have you ever tried to follow five hundred nested for loop and if statements? Shit is nuts...

  • source
  • [–] 1 point 1 hour ago

    repeat that for like one hundred billion times and now THAT's innovation

  • source
  • [–] 42 points 6 hours ago (1 child)

    You don't have to be an AI developer to not understand your code. I not understand my code every day.

  • source
  • hideshow 1 child comment
  • [–] 70 points 7 hours ago (5 children)

    The human brain is far too complex to understand

    looks inside

    it's just meat

  • source
  • hideshow 5 child comments
  • [–] 4 points 3 hours ago

    If statements are some of the largest sources of bugs in software.

  • source
  • [–] 129 points 9 hours ago (10 children)

    More accurately:
    >looks inside
    >matrix multiplication

  • source
  • hideshow 10 child comments
  • [–] [S] 9 points 5 hours ago (4 children)

    Yes, it is inherently incorrect, I just think that increases the fun factor. It reminds me of that guy who wrote a prime-checker for his thesis and it was just a massive list of if-statement

  • source
  • parent
  • hideshow 4 child comments
  • [–] 2 points 5 hours ago (2 children)

    A transistor is on or off. It's all if-then.

  • source
  • parent
  • hideshow 2 child comments
  • [–] 20 points 8 hours ago (2 children)

    Max(input*weights, 0) is an if in a sense, I guess.

  • source
  • parent
  • hideshow 2 child comments
  • [–] 3 points 6 hours ago (1 child)

    Good old ReLU. I've heard that CNN's perform better when using ReLU as the activation func, and that's probably why; ReLU acts as a filter on the image's features.

  • source
  • parent
  • hideshow 1 child comment
  • [–] 3 points 6 hours ago

    Nah. Without a nonlinearity you just get a linear combination of inputs instead of the output of a deep neural network. ReLU or Ramp is just the simplest possible non linearity. Using a simple function can enable using deeper networks yielding even better performance.

    It's actually somewhat of a headache, numerically. Works well enough tho.

  • source
  • parent
  • [–] 44 points 8 hours ago (1 child)

    It’s just y = k*x + m a few trillion times.

  • source
  • hideshow 1 child comment
  • [–] 30 points 9 hours ago

    come on, it’s not that simple!

    there’s also linear algebra

  • source