Eight Minutes

  • Comments posted to this topic are about the item Eight Minutes

  • I have to admit that I'm seriously concerned about AI, especially but not limited to when it's been asked to do something in SQL.

    When it comes to such a thing, even with some "common tasks", it turns out to be little more than a "consensus engine".

    I've asked it to do several simple things that it has gotten wrong and, sometimes, dangerously so but with a example of code that works and never comes close to testing the "edge cases" for the "Silent Failures", which are incorrect answers that cause no error.

    I also didn't want people to think that I found just an anomaly that would be "fixed soon", kind of like the early PEMDAS and "Number of "R"s in "Strawberry" problems.  I've been doing some simple tests now for a couple of years and the answers have not changed even through several revisions of different AI engines.

    The surprising reason is that it's still just a "consensus engine" and the human consensus as to the correct answers is still incorrect.  I've only found one other person on the internet that understands both the problem and the correct answer.

    I'm slowly working on a couple of articles from some very common issues but my real point is, never ever trust code no matter who or what wrote it.  And, when you test, always make sure to "test the edge cases".

    --Jeff Moden


    RBAR is pronounced "ree-bar" and is a "Modenism" for Row-By-Agonizing-Row.
    First step towards the paradigm shift of writing Set Based code:
    ________Stop thinking about what you want to do to a ROW... think, instead, of what you want to do to a COLUMN.

    Change is inevitable... Change for the better is not.


    Helpful Links:
    How to post code problems
    How to Post Performance Problems
    Create a Tally Function (fnTally)

Viewing 2 posts - 1 through 2 (of 2 total)

You must be logged in to reply to this topic. Login to reply