• 0 posts
  • 31 comments
Joined 3 years ago
Cake day: August 2nd, 2023
  • I fundamentally cannot agree with that take. How do you fix something if you don’t know why the current thing doesn’t work?

    Is the interface obtuse?

    Are the controls too manually complex to operate?

    Is the tutorial instruction flat-out wrong?

    Are they talking out off their ass about something they heard on hearsay?

    Were they taught secondhand, and poorly, by someone else on how to operate Thing?

    Please don’t try to imprecisely apply soft inclusivity to technical problems. If someone only says the stairs are difficult for them, don’t just change them into a slide because you accepted there needs to be change. This isn’t about accomodating someone’s lifestyle choices, this is (positing) dropping/adopting a standard based on vague dissent.

  • This feels like a shitpost that idiots misappropriated as a real method.

    Why would you want a method based on a famously developmentally-challenged child? It isn’t “dogged perserverance despite setbacks”; it’s “he’s literally too stupid to evaluate his process and teach himself a better way to solve even the simplest tasks.”

  • Read the images top-to-bottom, left-to-right.

    They used a shortcut to call the LLM to fix an error. The final “fix” was for the model to add a comment after the error saying “this is the error”. It’s about at the level of a real developer leaving a comment to themself saying “FIXME [later]”, instead of just fixing it.

  • Then give them some code and tell them to explain it. Having testees regurgitate algorithm implementations from memory is just testing rote memory.

    I thought we had all lived through terrible public schools that just tested memory, and resolved to do better.