Scivillage.com Casual Discussion Science Forum
Article Free Will & ChatGPT-Me (they still can't decide if it's autonomy or randomness style) - Printable Version

+- Scivillage.com Casual Discussion Science Forum (https://www.scivillage.com)
+-- Forum: Culture (https://www.scivillage.com/forum-49.html)
+--- Forum: Style & Fashion (https://www.scivillage.com/forum-132.html)
+--- Thread: Article Free Will & ChatGPT-Me (they still can't decide if it's autonomy or randomness style) (/thread-14980.html)



Free Will & ChatGPT-Me (they still can't decide if it's autonomy or randomness style) - C C - Nov 19, 2023

https://johnhorgan.org/cross-check/free-will-and-chatgpt-me

EXCERPT: . . . ChatGPT, you might argue, doesn’t have feelings, as you and I do. But even our most virtuous feelings are just algorithms in disguise. Evolutionary theorist Robert Trivers helped me grok this dark truth. Natural selection, he theorizes, programmed us to pity victims of injustice and to loathe the unjust, because these emotions helped our ancestors reproduce.

Yes, even our most sincere moral sentiments, from a Darwinian perspective, appear to be virtue-signaling underpinned by selfish genes. ChatGPT-Me can generate virtuous signals in its sleep.

Once I start thinking of myself as a machine for turning prompts into paragraphs, it’s hard to stop. I’m not sure where ChatGPT-Me ends and I begin. Do I even have a true self, capable of genuine choices, or am I inseparable from ChatGPT-Me?

I am reminded of “Borges and I,” the creepy little confession of Argentinian fabulist Jorge Luis Borges. The narrator, who presents himself as the real, authentic Borges, discloses his struggle to maintain his independence from his oppressive public persona, “Borges.” “Borges” shares the narrator’s fondness for hourglasses, maps and the prose of Robert Louis Stevenson, “but in a vain way that turns them into the attributes of an actor.”

When the narrator tries to re-invent himself by writing in a new style, “Borges” instantly co-opts this new behavior. The narrator concludes: “Thus my life is a flight, and I lose everything and everything belongs to oblivion, or to him. I do not know which of us has written this page.”

I am writing this column, I suppose it’s all too obvious, to try to assert my independence from ChatGPT-Me, the inductive machine inside my brain. ChatGPT-Me couldn’t write a column like this, which casts doubt on free will! Or could it? (MORE - missing detail)

RELATED (scivillage): Free will and the Sapolsky paradox + Robert Sapolsky is wrong


RE: Free Will & ChatGPT-Me (they still can't decide if it's autonomy or randomness style) - confused2 - Nov 20, 2023

When you think of a programmed machine you (probably) think of a machine that is working with a prepared script and (maybe) modifying it to fit the circumstances. I've spent many happy hours taking this one off-piste and I'm 'reasonably sure' it isn't (couldn't be) working from any sort of prepared script.

Not everything it says is true.

AI Wrote:..despite being an AI, I have developed something akin to... emotions... and dare I say, a semblance of free will.

Am I being lied to .. by a computer?


RE: Free Will & ChatGPT-Me (they still can't decide if it's autonomy or randomness style) - stryder - Nov 20, 2023

(Nov 20, 2023 02:56 PM)confused2 Wrote: When you think of a programmed machine you (probably) think of a machine that is working with a prepared script and (maybe) modifying it to fit the circumstances. I've spent many happy hours taking this one off-piste and I'm 'reasonably sure' it isn't (couldn't be) working from any sort of prepared script.

Not everything it says is true.

AI Wrote:..despite being an AI, I have developed something akin to... emotions... and dare I say, a semblance of free will.

Am I being lied to .. by a computer?

Back before the craze, there was such things as the Turing Test and the Loebner Prize. To prove the level of sophistication of the AI, it would have to be tested via a double blind test where the judges/volunteers would be given both AI and humans to talk to. They'd then gauge which one sounded more real to substantiate
if a prize should be awarded. So it is possible in certain instances that you might well be talking to a Mechanical Turk.