1

How essential is transparency in the AI future? Not just as a safeguard against anti democratic tendencies, but whether it is innately abuse to train AI to pose as a human, without clearly signallying you are using AI: in exams, otherwise human interaction, and anything really.

This post was not AI generated.

Improve this question
3
  • when is a lack of disclosure abuse? i have no idea
    – user73019
    Commented Mar 10 at 16:29
  • 2
    As is, the question is very opinion-based: essential to whom or to what? abuse under which ethical norms? You could either try to specify those or ask about references of philosophers who discussed this question in specific settings.
    – Johan
    Commented Mar 11 at 1:30
  • Is this a question that philosophy can answer? Presuming it can, how would philosophy answer it? Via logic? Via scientific investigation? The latter would require agreement as to objective. If we all agree on what we're trying to accomplish, it becomes a pragmatic issue as how best to do it. If it is a logical question, then no agreed upon objectives/values are needed. Afterall, we all agree w/regard to mathematical truths. No utilitarian questions there.
    – Gerry
    Commented Mar 31 at 1:33

1 Answer 1

Reset to default
-3

  1. One can pose this question to ChatGPT.

    I stated the question in different forms, even questioning that transparency is essential at all.

    I always got a similar list of arguments which emphasize the value and the necessity of transparency. The answers were clear, structured and listed a set of several criteria in favour of transparency. I assume, that most of us would sign onto this list. And we would notice that ChatGPT presents two or three additional criteria which oneself did not think of.

    One even gets good proposals when asking how ChatGPT does ensure to follow the principles of transparency.

  2. My conclusion from experiences with ChatGPT like these:

    At the present, one cannot rely that the developers of ChatGPT actually follow the principles of transparency. In general, the community does not have access to the algorithms and the training data. Even worse, the general user has the competence to understand and to assess the algorithms.

    Therefore, each text, which has been generated by AI should be marked correspondingly. Then the user has to make his decision about the proposed answers and the validity of the presented arguments, like in any discussion.

Improve this answer
5
  • 1
    dunno why this was downovoted, as it seems sensible enough, if sub fun. personally i feel it would likely be better if this stuff was just public access, rather than giving the illusion of control and transparency. include some kind of water stamp with it, who cares
    – user73019
    Commented Mar 10 at 18:34
  • dyt you can fall in love with an ai bot or just care about it as if you were :(
    – user73019
    Commented Mar 10 at 19:26
  • 1
    Sorry, but I cannot decode all your abbreviations because I'm not an English native speaker.
    – Jo Wehler
    Commented Mar 10 at 19:33
  • 1
    oh ok sorry, i hope i caused no offence. "do you think" is what i was getting at. anonymity is so strange...
    – user73019
    Commented Mar 10 at 19:33
  • 1
    OK, no abbreviations :-) But what do you mean?
    – Jo Wehler
    Commented Mar 10 at 19:43

You must log in to answer this question.