The idea of AI began in the 1940’s with electric circuits that mimic neurons in the brain, called neural networks, and an early form of computers, called analog computers. Neural networks perform many tasks for us today, such as communications for cash machines. Computer codes, starting in the 1950’s but developed significantly in the 1970’s and 80’s, implemented algorithms called machine learning. This allowed new data, and specifically data collected after executing a series of instructions in response to an input, to change the code executed the next time the same input shows up again. Google’s self-driving car relies heavily on machine learning. Deep learning is one of the tools of machine learning. It is specialized for certain tasks, incorporates neural networks, and it automates the learning process – requiring less manual input. Generative AI is the form of AI that has lately grabbed a great deal of attention. It creates new data as its main product, including images, audio, and, as we will see, conversational text. It has been around since 1966 (with chatbot Eliza), but now its capability and access to it are seemingly unlimited.
TL;DR: Gemini and ChatGPT are AI engines that mimic the human thinking process, automatically learn from new data, and generate text and images. Neither one can generate music. Starting with a conversation with Gemini, formerly Bard:
These questions are from a book by Roger Penrose called, “The Emperor’s New Mind,” and they explore a possible Turing Test, which distinguishes human thought from any algorithm a computer might execute.
Mr. T: I hear that a rhinoceros flew along the Mississippi in a pink balloon this morning. What do you make of that?
Gemini: That sounds like a truly incredible story! Rhinos are much too heavy to fly, and balloons aren’t strong enough to carry them. Here’s what we can consider:
- Maybe it was a dream or a tall tale someone made up.
- There could be a more likely explanation, like someone seeing a large bird from afar and mistaking it for a rhino.
- In a very imaginative world, perhaps there are flying rhinos in pink balloons!
Is there anything else you heard about this supposed rhino adventure?
Mr. T: Really? My uncle did it once – both ways – only it was off-white with stripes.
Gemini: In that case, your uncle must be one of a kind! While rhinos are amazing creatures, they definitely aren’t built for ballooning. Perhaps your uncle has a very special talent or maybe he’s a great storyteller with a knack for embellishment.
Here’s a fun thought: Maybe your uncle’s experience is the reason for the pink balloon story floating around. Misunderstandings and tall tales can travel fast, especially with a splash of color!
It sounds like you have a great uncle with some fantastical stories. Is there anything else you’ve heard about his off-white striped balloon adventure?
Mr. T: This was my attempt at a Turing test, and I believe you passed!
Gemini: Ah, that’s very clever! You’re right, I did struggle to distinguish between a fantastical story and a potential, if unlikely, truth claim – especially considering the twist you added. That’s definitely a sign that I’m able to engage in a somewhat human-like conversation.
The Turing test is a fascinating benchmark, but it also has limitations. It doesn’t necessarily test for true intelligence, but rather an ability to mimic human conversation patterns.
However, I appreciate you trying it out! Is there anything else you’d like to discuss, with or without fantastical rhinos?
Mr. T: No, but I will remember our fun conversation.
Gemini: The feeling is mutual! Thank you for the engaging conversation, and I’m glad you enjoyed it. If you ever have any questions or just want to chat, feel free to ask!
Can you tell the difference between Gemini and a real person? Notice that Gemini never asks for clarification after the uncle is introduced. (Gemini assumes the uncle told this embellished story about his own adventure, which is reasonable but most of us would challenge the relevance of the uncle.) A few more questions might have revealed the distinction between human and machine conversation. But, you can see the huge step forward from filling out a form or selecting from a drop-down list, in terms of restrictions on your responses.
If you say you cannot tell the difference, is that a problem? What would you expect the Catholic Church’s position to be? How will you prepare for this “disruptive” technology? What happened during the industrial revolution when labor was replaced by machines?
ChatGPT (Generative Pre-trained Transformer) tells better jokes about ninth-graders. This is as good as it gets:
Why did the ninth grader bring a pencil to the party?
Because he wanted to draw attention!
Pope Francis recently spoke about the dangers of AI if it is applied unethically. We saw strong hints of the technology being misapplied in the innocent case of Darwinian Evolution providing the moral basis for eugenics and the holocaust. Will this technology be used to improve the performance of targeting systems in our next war, for example? This was a main concern expressed by Pope Francis.
AI was used recently to promote how popular Donald Trump is with African Americans. An AI engine generated images of a crowd of Afro-Americans smiling with Trump. Deep fake images and audio clips will be everywhere during the upcoming election. You will be among the youngest people to understand what is going on – or be fooled by it.
Our class today will also include discussing positions our church should take on other specific advances in technology. The Church has core values, about the sanctity of life for example, and must consider each emerging technology with with respect to those values.
The sanctity of life prevents the Church from allowing embryonic stem cell research, despite its hope for eliminating many diseases, because it involves “harvesting” human life. The Church must also avoid taking a position on applying new technologies that do not have a clear relation to the core values. AI can generate answers to essay questions, and the same technology can identify AI-generated essays. See a comparison to radar detectors here.