Con without consciousness

It has been said, "there is no new thing under the sun." (Ecclesiastes 1:9)

Readers who were on the planet in 1966 might remember ELIZA, an early chatbot (anachronistically existing before that designation existed).

Consciousness (described here in Wikipedia) is not one of its attributes; ELIZA is clearly a mechanism.

The source code for this mechanism was uncovered/discovered by software archeologists in 2021.

The author is somewhat jaded about all of this AI hype, having decided against pursuing a PhD in the field of artificial intelligence—around 1986 (forty long years ago)—largely because of having stumbled upon a brief verse in the Doctrine and Covenants which asserts that "Intelligence ... was not created or made, neither indeed can [it] be."

In his view, this means that the term AI is an oxymoron: anything claiming to be AI is either not intelligent or is of natural origin. The former possibility is as difficult to test as any assertion about relative intelligence; the latter is somewhat chilling.

The Turing test (which ought to be widely understood) places ELIZA on the spectrum—with respect to chatbots—between GPT-3.5 and GPT-4.5 (the latter of which is considered to have passed the Turing test).

Without there actually being consciousness present, AI is a con to which we humans seem to succumb too easily.

Notes

The term "chatbot" was first used sometime in the 1990's, apparently a shortened form of "chatterbot" which appeared in the title of a paper published in 1964.

The exact scriptural reference is D&C 93:29, "Man was also in the beginning with God. Intelligence, or the light of truth, was not created or made, neither indeed can be." To the author, this implies that when God creates a human spirit, a pre-existing intelligence somehow enters into it or combines with it; a newly minted spirit somehow embodies the intelligence.

By "somewhat chilling" the author means to imply that somehow a sufficiently complex capability in a computer program might attract a pre-existing intelligence to enter into it, endowing it with intelligence. To think that humans could do this would be hubris on the order of the tower of Babel.

GPT-3.5 and GPT-4.5 designate Large Language Models (LLMs) which, when wrapped with an interactive shell, allow for humans to chat with them.

The Turing test for intelligence in machines was proposed (by Alan Turing (of The Imitation Game (2014) fame)) way back in 1950.

The author did not use AI to assist in writing this article—despite the presence of em dashes (each introduced manually by the author using his own fingers (to combine the three keys Shift-option-hyphen)). He confesses to having sometimes taken information directly from Google search's prominently placed AI-generated content at the top of search results (rather that scrolling down and clinking on an actual link (which he strives (but often fails) to always do)).