ARTIFICIAL INTELLIGENCE (AI) AND GOD

The July/August issue of Christianity Today magazine is mostly devoted to essays involving issues brought up by the rise in popularity of artificial intelligence. That prompted me to ask Google AI a number of random theological questions just to see what it would spit out. The results were as expected, as you can read below.

Q. Does man have free will?

    A. “The concept of free will is a complex and debated topic.”

Q. Was Jesus divine?

    A. Yes, within the Christian faith, Jesus is believed to be divine.”

Q. Is it true that Jesus existed?

    A. “Yes, historical evidence suggests that Jesus of Nazareth was a real person.”

Q. Is there life after death?

    A. “The concept of life after death is a fundamental question explored across cultures and religions, with no single definition.”

Q. How did life on earth begin?

    A. “The beginning of life on earth is a profound mystery, but current scientific understanding suggests...”

Q. How could the universe start from nothing?

    A. “The universe's origin from 'nothing' is a complex topic that challenges our everyday understanding.”

Q. Are all Roman Catholics going to hell”

    A. “Based on Catholic teachings, the answer is not all Catholics are going to hell.”

Q. Who wrote the book of 1 Timothy?

    A. “The Apostle Paul is traditionally considered the author?”

There was a little more to each AI reply, but the opening words I have quoted will give you a good idea of what the rest was like. As you can see, the not-quite-yet omnipotent AI generally provided wishy-washy replies opting out of making any judgment call. And in other cases, it turned the question over to what he/she/it felt was the best authority to answer it.

Thus, in questioning it on a matter involving Catholicism, AI let Catholic beliefs determine the best reply. Similarly, when asked about Jesus' divinity, it let the consensus within Christianity answer. And then it turned over the answer to “tradition” when asked about a question of biblical authorship. In all of these instances, it is sort of like a judge asking a suspected felon if he is guilty or innocent and then abiding by that person's answer.

In fact, there were only two of the queries above that were answered with any sort of assurance. And even then, not all people would be satisfied with the reply since both relied on one's confidence in the particular sources of the authority. I am speaking about what historians feel regarding the historical Jesus and what current science has to say regarding the beginning of life.

I don't know if these AI responses are an improvement or not on the old pre-AI opening replies to questions on Google. For example, when asked, “Is there a God?” pre-AI Google still states, “Theism, the proposition that God exists, is the dominant view among philosophers of religion. In a 2020 PhilPapers survey, 69.50% of philosophers of religion stated that they accept or lean towards theism, while 19.86% stated they accept or lean towards atheism.” At least this reply cites its original source, even though (1) there is still reliance on a biased and/or atypical population and (2) it commits the well known fallacy of counting evidence instead of weighing it.

It was this latter fault that gave rise to the “Authorized Version” of the King James Bible by which scholars at the time determined which of the variations in the available Greek manuscripts to use as the basis of their translation. They adopted the “Majority Text,” meaning that they counted all the available manuscripts at the time which read a particular way for each verse and went by the majority rule. By contrast, today's translators tend to assign much greater importance to the earliest manuscripts due to the known tendency for more and more errors in copying to appear as a document is recopied and recopied.

I realize that Google's current AI is still in its infancy, but all such attempts tend to discourage those who are interested in really learning more from going back to the original sources in order to (1) determine the scholarly background of the various authors behind the “majority view,” (2) look for any possible biases those authors may have, (3) and to investigate and weigh the reasons they give for their conclusion. So much of what AI spits out is also highly dependent on which algorithm the computer programmers happen to adopt, and that algorithm is rarely made public for anyone to independently assess its validity to adequately address difficult questions unless they involve black-and-white areas that can be easily quantified.

It is well known that the “majority view” on practically any subject, particularly theology, has changed drastically just in the last hundred years or so. Liberal biblical scholarship coming out of the mainstream denominational seminaries around 1900 would overwhelmingly favor one set of answers to the questions above. But since that time, more conservative Bible scholars are in the ascendant in publishing their views and would give a much different set of answers to those same questions. Thus, the majority view really is meaningless to talk about, since it is so dependent on which population of “experts” you choose to poll and which ones you chose to exclude. That would take a value judgment, something for which no one person or committee is qualified to do, and certainly not an artificial intelligence machine, at least not for the foreseeable future.

Since AI is such a current subject of interest, I will write two more short posts attempting to share my own opinions and experience on two related issues: AI and human beings and AI as a replacement for God.