3 ms·

Since AI is the last great problem in computer science, I can't help but feel a tremendous sense of sadness and foreboding about this. I always wanted to dabbl

Since AI is the last great problem in computer science, I can't help but feel a tremendous sense of sadness and foreboding about this. I always wanted to dabble in AI in that far off mythical future when I have the time to do so. But if it all gets patented, what's the point? Might as well patent relativity, or quantum mechanics.

deelowe 8y ago
There's still quantum computing.
dannypgh 8y ago
So, does P=NP? There's a lot of great unsolved problems in most fields, including CS.
- panopticon 8y ago
  No, see... humans won't be doing any more problem solving when we have AGI. We'll just kick it and drink Mai Tais on the beach while they do all the work.
  - john_moscow 8y ago
    I wish... I'm afraid it's more like being caught in endless infighting while trying to figure out who's more important. Look at any big corporation...
bvc35 8y ago
AGI is achievable, but we are along way away from it. There's still plenty of time to get involved! If you can't get involved now, at least try to influence lawmakers to remove the privilege of firms to patent algorithms, which will stifle research progress.
therealtomsmith 8y ago
I love when people say "dabble in ai". The mathematical foundation to even begin truly understanding ai isn't something you can dabble your way through. - so am i being downvoted because im wrong or because i hurt your feelings? jesus fucking christ.
- zackmorris 8y ago
  I upvoted you a little because I think the perception that the mathematical foundations of AI are too complex is valid. AI is actually pretty straightforward from what I can tell, but it has a problem with notation. I think that probability and quantum mechanics also suffer from (context-heavy?) notation and steep learning curves. I'm hopeful though that as more people dabble in AI that the notation problems will be alleviated somewhat. My gut feeling is that its abstractions are still too low-level, like the early days of CS when people worked with binary, assembly or even LISP. Once we begin to see the forest for the trees (something analogous to MATLAB or Elixir but for AI), then I think we'll finally see the utility of AI realized for mainstream developers.
- gimmeThaBeet 8y ago
  Probably it reads like textbook gatekeeping and doesn't add much to the discussion. Granted this doesn't really either, but I'd rather people be interested in learning something rather than just be intimidated because they can't "really" learn anything about AI. That's just plain mean-spirited.

deelowe 8y ago

There's still quantum computing.

dannypgh 8y ago

So, does P=NP? There's a lot of great unsolved problems in most fields, including CS.

panopticon 8y ago
No, see... humans won't be doing any more problem solving when we have AGI. We'll just kick it and drink Mai Tais on the beach while they do all the work.
- john_moscow 8y ago
  I wish... I'm afraid it's more like being caught in endless infighting while trying to figure out who's more important. Look at any big corporation...

bvc35 8y ago

AGI is achievable, but we are along way away from it. There's still plenty of time to get involved! If you can't get involved now, at least try to influence lawmakers to remove the privilege of firms to patent algorithms, which will stifle research progress.

therealtomsmith 8y ago

I love when people say "dabble in ai". The mathematical foundation to even begin truly understanding ai isn't something you can dabble your way through. - so am i being downvoted because im wrong or because i hurt your feelings? jesus fucking christ.

zackmorris 8y ago
I upvoted you a little because I think the perception that the mathematical foundations of AI are too complex is valid. AI is actually pretty straightforward from what I can tell, but it has a problem with notation. I think that probability and quantum mechanics also suffer from (context-heavy?) notation and steep learning curves. I'm hopeful though that as more people dabble in AI that the notation problems will be alleviated somewhat. My gut feeling is that its abstractions are still too low-level, like the early days of CS when people worked with binary, assembly or even LISP. Once we begin to see the forest for the trees (something analogous to MATLAB or Elixir but for AI), then I think we'll finally see the utility of AI realized for mainstream developers.
gimmeThaBeet 8y ago
Probably it reads like textbook gatekeeping and doesn't add much to the discussion. Granted this doesn't really either, but I'd rather people be interested in learning something rather than just be intimidated because they can't "really" learn anything about AI. That's just plain mean-spirited.