←back to thread

301 points lukeio | 1 comments | | HN request time: 0.246s | source
1. godelski ◴[] No.46239433[source]

  > Craft software that makes people feel something
Meta, Google, and all of FAANG already did that. They crafted software that made people feel hate, anger, depression, but sometimes joy. It's nice to get those cute animal posts when doom scrolling. It's a nice break from "you're all going to die", "everyone is dumb except you", and "you're powerless".

Joking aside, I do very much agree with the OP. But I also wanted to note how things can get perverted. Few people are actually evil and most evil people get there slowly. What's that old cliché that everyone forgets? "The road to hell is paved with good intentions". The point is to constantly check that you're on the right path and realize that most evil is caused in the pursuit of good because good is difficult to do.

But also I wanted to share a Knuth quote

  | In fact what I would like to see is thousands of computer scientists let loose to do whatever they want. That’s what really advances the field
  - Donald Knuth
I am fully with him on this. It is the same reason Bell Labs had so much success.

  How do you manage a bunch of geniuses?[0]. You don't. 
You let experts explore. They already know the best ways forward. Many will fail, but that's okay. In CS one of the biggest problems we have is that we try to optimize everything, but we're also really bad at math. If you want to optimize search over a large solution space with a highly parallelized processor you create a distribution. It's good to have that mean but you need the tails too and that's what we lose. You tighten the distribution when you need focus on a specific direction but then relax it to go back to exploration. But what we do, is we like railroads. We like to try to carry all the groceries from the car in one trip. We like to go fast, but don't really care where. We love to misquote Knuth about premature optimization to justify our laziness and ignore his quotes about being detail oriented and refining solutions.

I think progress has slowed down. And I think it's because we stopped exploring. We're afraid to do anything radical, and that's a shame

[0] Knuth has another quote about programmers not being geniuses lol