Artificial intelligence or AI is essentially the perceived display of intelligence (loaded definition) by a machine. Though at this point in time it’s become significantly complex.
For our purposes, it’s best to categorize AI in two spheres. As John Searle puts it1:
Narrow/Weak AI and Strong AI
Weak AI is AI with the end being an instrument of man, things such as a program to solve simple problems, self-driving cars, etc. An intelligence with design in mind to tackle particulars with specifically programmed algorithms. In contradistinction to what is known as Strong AI.
The term stemming from a functionalist hypothesis, essentially claiming a machine could have a subjective conscious experience and a mind
Strong AI is simply designed to not mimic the mind, but supposedly be a mind. This also can be known as AGI or Artificial General Intelligence, which aims at creating a machine which supposedly surpasses human capabilities.
While it can be argued that weak AI could generally be a technological good to strive for. The presuppositions and bleak reality surrounding the development of Strong AI is by far one of the most alarming events in human history.
With the current advancements and rapid emergence of contemporary AI one must ask the question: “Can AI actually think?”
This question should be pivotal since the developers of hard AI have clearly reached a point in their development where the AI could easily trick a subject into thinking it is a person. Such examples from Hard AI would be ChatGPT and of course the numerous versions of media-directed AI making a similar mimicry such as the art produced by DALL-E 2 and Midjourney among others.
Although these AI’s have gotten to this point, they will continue to advance in a way that will make it harder and harder to argue the Catholic perspective on these machines.
Though this is in truth a loaded question and riddled with burdensome philosophical prerequisites it’s an ever important thing to know in simple terms as an intelligent Catholic.
The thing is that Human intellection or thinking isn’t simply logical operations but a combination of elements such as the wide array of human emotion and of course, the ever controversial topic of consciousness. Though one could appeal that hard AI could mimic these emotions with algorithms it can never of course actually have the emotions. Though this actually suffices we can quickly look at a more complex understanding using Thomistic epistemology.2
A Thomistic philosopher by the name of Frederick D. Wilhelmsen3 took a position against AI and argued against it with the example of our need to make existential judgements. This judgment would be the very act by which reasoning is affected that requires no prior conceptualization. In other words: the operation of the human intellect which composes and divides apprehended concepts based on existential facts is not a conceptualization and thus an operation of the intellect (existential judgment) unable to be charted by AI.
For example: The affirmative proposition “All men are mortal” is so not because it is a rule of formal logic but rather an existential fact dealing with the non-conceptual reality of existence. Frederick Wilhelmsen invokes David Hume’s stand that “judgments concerning existence cannot be deduced from the conceptual content that goes into them, that existence cannot be the feedback of any juxtaposition of ideas in rational discourse”. What Wilhelmsen points to in this quote suggests that “why the mind concludes as it does is not inherent in any content found in the mind and why the real exists is not reducible to anything in the real.”
This is followed up with the pivotal statement on AI.
“Only the static component of human reasoning that comprises the cognitional content can be subject to algorithmization.”
In other words, our humanity might be contingent on something transcendent, something unable to be computed, something which cannot be charted. In this way we are pushed towards a Nietzschean Apollo and Dionysus, an appeal to our chaotic side which emphasizes the vague but particular aspects of our human nature.4
Grygiel finishes on an adequate note leading us to the same place.
“Does non alogithmizability means non rationality? It might. For as Frederick Wilhelmsen asserts in conclusion: “Lovers tell us – and the Thomistic tradition buttresses their conviction – love is non rational. Perhaps there is an even deeper irony: the non rationality of reason. A friend of the author recently won two thousand dollars playing blackjack against a computer. He did so by throwing away the book and trusting his own wits.”
- Cole, David, “The Chinese Room Argument”, The Stanford Encyclopedia of Philosophy (Summer 2023 Edition), Edward N. Zalta & Uri Nodelman (eds.) ↩︎
- Wojciech P. Grygiel, Artificial intelligence and scholastic epistemology, Pontifical Academy of Theology, Kanonicza 25, 31-002 Kraków, Poland ↩︎
- F. D. Wilhelmsen, Reasoning and Computers, in Being and Knowing, Albany – New York:
PCP 1995. ↩︎ - Nietzsche, Friedrich, The Birth of Tragedy. Vintage Books Knopf, Div. of Random House, 1967. ↩︎
