• 0 Posts
  • 287 Comments
Joined 2 years ago
cake
Cake day: April 6th, 2024

help-circle


  • In my experience there are three ways to be successful with this tool:

    • write something that already exists so it doesn’t need to think
    • do all the thinking for it upfront (hello waterfall development)
    • work in very small iterations that doesn’t require any leaps of logic. Don’t reprompt when it gets something wrong, instead reshape the code so it can only get it right

    The issue with debugging is that it doesn’t actually think. LLMs pattern match to a chain of thought based on signals, not reasoning. For it to debug you need good signals in your code that explicitly tell what it is doing and the LLMs do not write code with that level of observability by default.

    Edit: one of my workflows that I had success with is as follows:

    • write a gherkin feature file describing desired functionality, maybe have the LLM create multiple scenarios after I defined one to copy from
    • tell the LLM to write tests using those feature files, does an okay job but needs help making tests run in parallel.
    • if the feature is simple, ask the LLM to make a plan and review it
    • if the feature is complex then stub out the implementation in code and add TODOs, then direct the LLM to plan. Giving explicit goals in the code itself reduces token consumption and yield better plans










  • zbyte64@awful.systemstoLefty Memes@lemmy.dbzer0.comMamdisappointing
    link
    fedilink
    English
    arrow-up
    0
    arrow-down
    1
    ·
    5 months ago

    “We want him to act like a hegemon and purge people based on ideology instead of what they have done.”

    Go read “Leadership and the Rise of Great Powers”, written by a member of the CCP to better understand what kind of leader Mamdani is vs the kind you are asking for and the consequences there of.





  • The part about robots doing backflips causes the robot to wear down faster has me thinking the whole “replace humans with humanoids” should be framed as comparative advantage rather than how many robots would be required to build itself. Given the number of humanoids required to replicate itself, you could take those same complex parts, rearrange them into non-humanoid configurations and have more output both in an interval of time and over the life time of those parts.