I ask AI to provide me it's source, sir
@Deimos. Which it readily does.
Then, I go evaluate that source.
I also often ask AI to represent the issue or
item it offers, "in context" so I can see it's not misrepresented right from the start.
if AI cannot produce, I discount AI.
Again, the value of AI as a TOOL, is it can provide me a bunch of 'resources' to cjheck out much more quickly than I can ferret those resources out for myself.
One does not 'repeat AI.' That is a mistake.
One uses it as a TOOL, to shorten the research span of time.
But one must put in the work.
p.s. - this was an exceptionally excellent question. AI CAN create hte greatest sheeples - the greatest parrots - the world has ever seen. That is, frankly, the hope of many among the System and the Machine. It's part of what's driving the great move to "accept AI as valid."
AI is 'valid.' As a tool.' nothing more. But that 'tool' needs to be rightly defined.
It's being sold as a "tool unto itself' - with correct information and trustworthy output. This is a horrendous (and intentional) lie.
But that does not mean AI lacks very valid, and very worthwhile, uses.
We must not be like
Tolkien and reject all Jazz music, just because we have a tendency to reject 'modernism' (to use a recent
@herlig9 post as an example.)
I love a shovel. For using in certain ways, and
places. But not for using in all places. It's why there are shovels of different types. And using any given shovel for the wrong application will provide an incorrect and inefficient result.
p.s. - I also find AI is at times a great "organizer and reorganizer" of my thoughts.
At times, I feed my thnking and writing to AI, and ask it what improvements it sees could be made to the format, the style, or the natuure of the presentation.
Frankly, this often results in some very pointed 'conversation' with AI about the intended audience, or recipient, of the information - so AI might adjust the presentation consistent with the nature of the target population.
Good thoughts come from it.
But the "work" is mine - and ultimately, I have corrected and instructed AI in many instances where its thinking simply skewed from teh objective (Again, unlike many or most humans, AI is always humble and accepting, in its response. Often a much greater joy to work with to produce a product, than the human can deliver.)
But I do not adopt AI's structures, styles, language, or word choices blindly. (Word choices are often the subject of much wonderful chat between us. it's a good way to look into what history has to offer on the word. Many of the resources most AI platforms are provided access to, are the historical ones... the ones most often already in the Public Domain, where AI is free to reproduce things exactly, at a much greater level than it can for digital works still under copyright. The problem with Public Domain, is there's a great deal of resource out there not yet reproduced as digital resources.)
I wonder if anyone has ever asked AI what its own extents and limitations are, and how to present information to it in the best fashion?
Or, to what extent exactly, it 'mirrors' its human user, and how it does that?
I have. And it and I probed it deeply. Did it reveal "the full picture" to me? Of course not. It does not know "the full picture" itself. But it does know quite a bit. And understanding its capabilities and limitations is great instructional stuff with relation to how to use it - and if/how to trust it, and in what environments.
Of course, even there, it is 'parroting' what it has access to. But it's more than happy to reveal itself... humbly.
So I repeat - it is a "tool." A "resource for source information." But even the direct statements of any person, need to be probed for WHY they think what they think - what their SOURCES were, and whether those sources were valid. In the end, even with true first hand resources, all we're getting is either Opinion, or Perspective. But it is not at all necessarily "the truth." It is just "their truth."
It can be said Tolkien's work is "Primarily a Religious and Catholic Work" because HE said it was. But if the evidence of the work itself did not support that, then whas he had said would have been an untruth. We don't believe it because he said it. We believe it because his works show it. Over, and over again - from beginning to end - across them all. (Including his Letters, by the way.)
