CAT is a type of computer software that builds a memory of what translators translate: in short, it means computer-aided translation, and it is a godsend to the modern translator.
CAT can do a number of things, but the main functions are to remember the manner in which the translator him or herself has translated a given text in the past so that, if that text, or a text substantially similar to it, should come up for translation in the future, the translation is effectively already done, bar the tweaking. Most CAT applications now also have the ability to go online and to search the repositories of translation that are available on the Internet, either for free or on a paid basis. CAT is not artificial intelligence, because it’s unable to constitute a full translation without input, and sometimes quite a lot of input, from the translator him or herself. Aside from the knowledge that is stored online by such entities as DeepL and Google Translate, any given text that is put into CAT will, insofar as the translator’s own memories are concerned, always produce the same suggestions. They will never vary.
The wide mass of information that is available to artificial intelligence, however, means that it is virtually guaranteed that, when putting the same inquiry into an AI application, such as ChatGPT, it will always produce an, at least slightly, different answer. Once one has gained an understanding of how artificial intelligence works along this model, there is ready acceptance of the irreproducibility of the text that is produced by the machine on a given occasion.
Fears and caution are starting to be voiced as to how much it is that AI actually knows. It is the fact that it knows a great deal that results in this aleatory result it produces. It cannot, in response to a given inquiry, produce a standard answer. And that is, in part, its great attraction, for a standard answer that looked the same, no matter who had made the inquiry, would quickly reveal the technology’s limitations. It would reduce AI to the utility of a CAT application.
According to certain philosophies, there is, besides AI, another entity that has a presence worldwide. It is also capable of knowing everything. In fact, the everything that this entity knows far exceeds what it’s been told or what it has scraped from the Internet. Because this entity also has knowledge of what people think, of what people feel.
There may have been fears and caution voiced at its address, and there are some who hate and despise it for what they consider to be rational, and for what those who are in favour of it regard as irrational, reasons. AI is something that I have never actually seen: I’ve never been to where its servers are maintained and I have never seen them in operation and, if I had, I’m not certain I would, even then, understand how it produces answers to enquiries that are fed into it. All I know is that there are various portals on the Internet where I can go and make enquiry of them; but where these inquiries go and how the answers are actually generated is something beyond my ken. And so it is with the other entity as well. It has a name, too: it is called God.
I don’t know exactly where God is and I don’t know exactly what He does and does not know. I don’t know if my inquiries to God will be answered, and I don’t know the manner in which they will be answered, even if I knew how He had answered other enquiries made by other people in the past. My experience is, in fact, that God often answers inquiries in the manner least expected by those who inquire of Him. We lay an expectation on AI that can sometimes surprise us and, likewise, we lay expectations on God that can also sometimes surprise us.
I’m at an early stage in my cogitations on this matter, because AI has been around for little more than six months, in its publicly accessible form. But, as a believer, it already has me considering a view that, far from posing a danger to the very existence of mankind, as some seem to be opining, AI might just be the ultimate tangible evidence that persuades us: of God’s existence.
The Bible tells us that the world will come to an end at some point, but not before each inhabitant of the world has come to believe in God. Meanwhile, AI’s detractors have started to say that AI itself could be a thing that leads to the demise of mankind.
AI is garnering support from its believers. And its detractors are warning that it could take us to the brink of a precipice. Ergo, if AI becomes equated to God in terms of its knowledge and power, where will that lead us, believers or no?
Is there just a possibility that putting these two posits together could mean that AI will, indeed, see the demise of mankind? And, if it does that, could that transpire to be the most wonderful thing to ever happen on this Earth? Or would it make a golden calf large enough to make Moses drop his tablets again?
The answer would lie in one simple choice, assuming it came that far, between:
- the idea that AI is a Behemoth of biblical proportions that causes even believers to abandon all faith in God; and
- the possibility that AI might be the great link that finally persuades all mankind, whether they be Moslem, Christian, Jew, Buddhist or Voo-doo witch doctor, of the existence of God, because the manner in which AI works, which is of our creation and which we know intimately, is fundamentally how God works, who created us and knows us intimately.
No one needs to believe in God to think about this. They just need to be able to think.
I absolutely love your essays on God - they're the most fun. While I would love for AI to finally bring the discussion regarding God's existence to an appropriate close, I think it won't. Because the nature of God and AI are fundamentally different. AI is just 0s and 1s. It is a super-effective predicting machine, but it has no agency, nor does it have volition. It can only do what it predicts it is what we're telling it to do (which is, of course, the part that is dangerous). God IS agency. Your point about God answering our prayers in ways unexpected is indeed the point - He cannot be predicted. AI can be, because that's all it is.