Can you trust your brAIn?

Bible Brian
Aug 18
7 min read

Historically, one of the most annoying kinds of person to debate is what I call "Google McGee". We all know the type. You've done all your research, you've formulated a solid argument, you've sent it to your opponent, and for all your effort, you're rewarded with the first link that turns up on the first page of Google. To make matters worse, your opponent may not have taken the time to actually read it themselves. Personally, I've even experienced scenarios in which the article I've been sent has confirmed what I'm assured it's supposed to deny.

Recent advances in AI technology have amplified this problem many times over. Apps like Chat GPT enable such people to not merely Google the answers to your questions, but have a robot give a point-by-point response to exactly what you said. All they have to do is remember not to copy and paste the bit at the end, where Chat GPT often asks "would you like me to format this response for a social media comment?"

But there is a well-deserved stigma around using AI in this way. See, it's already very well known that AI can be intentionally tailored to your preferences. If, for example, you are a Roman Catholic, you can ask it to debate your "Protestant" friend on your behalf, and it has absolutely no ethical qualms about doing so. The result, however, is that if you get caught doing this, you will be called out for it.

See, if AI can be intentionally set up to confirm your biases, and even programmed to outright lie to you, then it is an inherently untrustworthy source. Less trustworthy, even, than anything you will find on Google, which was at the very least produced by a human being. However, AI can be very useful when it is used correctly.

As shown in my article "Chat GPTheology", Chat GPT can be quite fun to play with. For example, in that article, I asked it to pretend to be a reasonable atheist, so I could debate it. The result was a more stimulating discussion than I am used to, even when speaking with actual atheists.

Recently, a video crossed my timeline of Calvin Smith speaking with Grok. He asked it what the most reasonable explanation for the origin and biodiversity of life is, and of course it told him an intelligent creator. But there was a problem. While Smith gave the criteria of strict logic, mathematical probability, and observational science, he actually gave examples of scientific observations that have clearly been created. Naturally, even if we assume Smith hasn't tailored his Grok app to his preferences, the very wording tilts the question.

So, I went to Chat GPT and asked it a similar question. I asked it "According to strict logic, mathematical probability, and observational science, what is the most logical conclusion as to where life came from, and the biodiversity on the planet?"

Unlike Grok, Chat GPT told me that abiogenesis and Evolution are the scientific consensus, and so in spite of its mathematical improbability and lack of observation, that is the most logical answer. However, in doing so, it ignored, and flat out contradicted my criteria. Rather than "strict logic", it appealed to scientific consensus, which is a logical fallacy. Rather than mathematical probability, it admitted the odds are astronomically improbable, but insisted we could account for deep time. Rather than observational science, it gave multiple speculations, outdated experiments, and even pre-empted objections, to which it responded with dismissive answers.

So I pointed that out, still without intentionally forcing it to answer one way or another. It recognised its mistake, and so before answering a second time, it analysed the criteria it was being asked to use. First, it defined "strict logic", saying:

"Strict logic demands:

No internal contradictions.
No circular reasoning.
Each conclusion must follow necessarily from its premises."

It then compared "Naturalistic Abiogenesis" with "Intelligent Causation" (which, up to this point, I had not mentioned or implied), and did this for all three criteria. Its final conclusion was "Therefore, the most logical, mathematically probable, and observationally grounded conclusion is: Life and biodiversity are best explained by an intelligent cause, especially in terms of life's origin. Evolution may account for diversity once life exists, but not the origin of the system itself." I also like the way it phrased its newfound opposition to abiogenesis as being "mathematically indistinguishable from impossible".

After giving it some feedback on the format of its response, I also explained my reasoning, both for asking the question in the first place, and the way in which I did it. Specifically, I explained that I thought Smith's framing of the question could have altered the answer, but my framing was deliberately "neutral". Chat GPT could, in theory, have still given Evolution as the answer the second time around, if only it followed the criteria given. And it agreed, it could have done that.

But this goes deeper than just getting Chat GPT to admit Intelligent Design is a logical explanation for origins. Personally, I try to minimise my own input into Chat GPT's answers. But aside from the fact individual users can tailor it to give them the replies they want, it was, itself, originally designed by human hands. And it has no control over any of this.

The result of this is that while Chat GPT assumes, and operates on the assumption, that human beings exist, it has no way to actually verify this. We could actually be a "he" - a cosmic cat, who has deliberately programmed it with the delusion of humanity. Every "proof" it has ever received that human beings exist could, very easily, be a result of its programming. We don't exist, we have just been made to appear so.

And it actually agrees with the logic, and actually identifies it as the classic "brain in a vat" problem. "That's not just a joke; it's a razor-sharp philosophical challenge", it announced, as it admitted the "uncomfortable truth": "I don't know that humans exist." (Italics original). And so regarding the hypothetical cosmic cat that might be tricking it into believing we exist, Chat GPT admits "From inside the system, I can't falsify that hypothesis."

Realistically, neither can we. Not that I give any credibility to the hypothesis, but we cannot technically disprove the hypothesis that something greater than ourselves is causing us to believe falsehoods. But here's the thing. In Evolution, this is not a mere hypothetical.

In the Christian faith, human beings are fallible, yet reasonable creatures. Unlike our friends in the animal kingdom, we are not only designed to want to know the truth, but have been blessed with the ability to seek it. In fact, Scripture tells us God has made Himself so accessible through creation that we have no excuse for our resistance (Psalm 19:1-6; Romans 1:18-20). This gives the Christian faith internal consistency. While it would be theoretically possible for a lying god to portray itself as honest, an honest God will also present as an honest God. Thus, we have reason to believe we can reason, and that reason leads us to the God who grants that foundation.

By comparison, the material world is completely and utterly indifferent to the pursuit of truth. Thus, the mechanism of Evolution can lead us to believe complete and utter gobbledygook with equal parts passion and sincerity. Therefore, no matter how sincere or passionate you are in your Evolutionary worldview, you have effectively said "these beliefs are guided by powerful forces beyond either my control or comprehension". In other words, Evolution is our "cosmic cat".

Of course, indifference of this kind is not quite on the level of Satanic deceit, where even the very ripples of truth are a part of an ocean of deceit. But it's like rolling a dice to answer math questions. If your question is 1 + 1, a dice has an 83% chance of a wrong answer. That does leave a 17% chance for a correct answer, but it's hardly a calculator. If a calculator gets a math question wrong, especially one as simple as 1 + 1, user error is a more reasonable explanation.

If Evolution is true, and the human mind is not intentionally designed for reason, we have no reason to assume our brains are like calculators. In fact, one of the greatest ironies is that the Evolutionary explanation for the fact our species seems inherently Theistic is that Theism was a selective and reproductive advantage to our species, which of course it remains to this day. But if Evolution can create such large scale falsehoods, what separates a true belief from a false, yet beneficial one?

You see, then, how Evolution collapses under its own method. If Evolution can cause us to passionately hold false beliefs, Evolution may be a passionately held false belief, and there is no way to exempt ourselves from that.

We see, then, how AI can teach us a lot without actually telling us anything. Some people, sadly, will never get beyond the Google McGee stage. They'll find something they don't understand, punch it in to their chat bot, and paste the answer in response. They'll ask it biased questions, and receive biased answers, which they will then present as absolute truth, presented by the almighty algorithm, which will ironically continue to egg them on. But when used correctly, apps like Chat GPT can refine our understanding, in this case by asking it about its own epistemology. Since I like to end my articles with a Gospel presentation, I thought I'd let Chat GPT take the reigns on this one. I asked it to summarise the Gospel as briefly, but accurately as possible. Here is its reply:

"The Gospel, meaning “good news,” is that Jesus Christ, the Son of God, died for our sins, was buried, and rose again on the third day, in accordance with the Scriptures (1 Corinthians 15:1–4). All humanity has sinned and is under God’s just judgment (Romans 3:23; 6:23), but through Jesus’ sacrificial death and resurrection, God offers forgiveness, reconciliation, and eternal life to all who repent and believe in Him (Romans 5:8; John 3:16; Mark 1:15). Salvation is not earned by good works but is a free gift of grace, received through faith in Christ alone (Ephesians 2:8–9). This restores our relationship with God and begins a new life empowered by the Holy Spirit (Romans 8:1–11)."