3 ms·

Until there is an AI that can take as input a JIRA ticket (potentially asking for clarifications) and can output a Pull Request, I'm not too worried.

nnoitra 4y ago
amag 4y ago
You misinterpret, I'm not worried about my job. I wouldn't mind working less but I'm worried that sooner than we think we will be reduced to pure consumers (we're already pretty far ahead in that respect) and then life will lose its meaning...
- toastmaster11 4y ago
  I think the human need to feel useful, or to produce something is more sociological than a base human need. When/if that becomes an issue, if social rejection was not a consequence of not being productive more people would be okay with simply participating in things for their own sake. For example, I don’t play Stardew Valley because I want to be the most elite virtual farmer.
  - 1-hour-ago 4y ago
    Maybe we're already seeing this. Seems like people would rather join a cult than not feel useful. Taking handouts from the man... er... machine and free time for hobbies probably isn't going to do it. Feeling useful could turn out to be part of basic human dignity.
    - OrsonSmelles 4y ago
      But those people have generally marinated their entire lives in a background frame that usefulness == worth and (as a generalization) are under-equipped to generate meaning internally! It's not obvious to me that, say, the fifth generation after human labor becomes superfluous will be more vulnerable on this axis than we are now when there's tension and uncertainty about it. (I'm sure plenty of people would happily jump in to question whether we'd make it five generations past that horizon, but that's not really the point I'm after here - let's assume we haven't Idiocracy'd/WALL-E'd ourselves to death.)
ilaksh 4y ago
The latest large language models can do that for some tickets. The only big thing missing is the ability to see and interpret application screens and maybe a memory limitation for many things. Considering how fast technology progresses we should all be worried and adjust our plans.
- nsxwolf 4y ago
  Tickets are rarely correct as written. Is the to AI going understand that and conduct meetings to hash out the missing details and corrections?
  - acuozzo 4y ago
    > Is the to AI going understand that and conduct meetings to hash out the missing details and corrections? Why not? ChatGPT is nearly there right now.
    - nsxwolf 4y ago
      Can ChatGPT have suspicions that a requirement isn’t quite right?
      - acuozzo 4y ago
        No, but I'd bet that it'll improve to the point of being able to determine ambiguous requirements and/or ones with insufficient detail. Given the popularity of "NoCode" I reckon that managers & potential clients are more willing to meet in the middle than one would otherwise assume. Since programming problems are recursively decomposable into sub-problems, all that would be required is to iterate on the ChatGPT-assisted requirements list until it's unambiguous and then repeat the process for each subproblem until the capabilities of AlphaCode are reached.
  - ilaksh 4y ago
    It can send a chat message and hold a dialogue. For some small tasks things that is enough. Again main limitation is really vision and memory but we have to anticipate there will be very significant progress on those things in the next few years.
jimbokun 4y ago
I don't see why that's an obstacle that can't be overcome in the very near future, given current rates of progress.

nnoitra 4y ago

amag 4y ago

You misinterpret, I'm not worried about my job. I wouldn't mind working less but I'm worried that sooner than we think we will be reduced to pure consumers (we're already pretty far ahead in that respect) and then life will lose its meaning...

toastmaster11 4y ago
I think the human need to feel useful, or to produce something is more sociological than a base human need. When/if that becomes an issue, if social rejection was not a consequence of not being productive more people would be okay with simply participating in things for their own sake. For example, I don’t play Stardew Valley because I want to be the most elite virtual farmer.
- 1-hour-ago 4y ago
  Maybe we're already seeing this. Seems like people would rather join a cult than not feel useful. Taking handouts from the man... er... machine and free time for hobbies probably isn't going to do it. Feeling useful could turn out to be part of basic human dignity.
  - OrsonSmelles 4y ago
    But those people have generally marinated their entire lives in a background frame that usefulness == worth and (as a generalization) are under-equipped to generate meaning internally! It's not obvious to me that, say, the fifth generation after human labor becomes superfluous will be more vulnerable on this axis than we are now when there's tension and uncertainty about it. (I'm sure plenty of people would happily jump in to question whether we'd make it five generations past that horizon, but that's not really the point I'm after here - let's assume we haven't Idiocracy'd/WALL-E'd ourselves to death.)

ilaksh 4y ago

The latest large language models can do that for some tickets. The only big thing missing is the ability to see and interpret application screens and maybe a memory limitation for many things. Considering how fast technology progresses we should all be worried and adjust our plans.

nsxwolf 4y ago
Tickets are rarely correct as written. Is the to AI going understand that and conduct meetings to hash out the missing details and corrections?
- acuozzo 4y ago
  > Is the to AI going understand that and conduct meetings to hash out the missing details and corrections? Why not? ChatGPT is nearly there right now.
  - nsxwolf 4y ago
    Can ChatGPT have suspicions that a requirement isn’t quite right?
    - acuozzo 4y ago
      No, but I'd bet that it'll improve to the point of being able to determine ambiguous requirements and/or ones with insufficient detail. Given the popularity of "NoCode" I reckon that managers & potential clients are more willing to meet in the middle than one would otherwise assume. Since programming problems are recursively decomposable into sub-problems, all that would be required is to iterate on the ChatGPT-assisted requirements list until it's unambiguous and then repeat the process for each subproblem until the capabilities of AlphaCode are reached.
- ilaksh 4y ago
  It can send a chat message and hold a dialogue. For some small tasks things that is enough. Again main limitation is really vision and memory but we have to anticipate there will be very significant progress on those things in the next few years.

jimbokun 4y ago

I don't see why that's an obstacle that can't be overcome in the very near future, given current rates of progress.