Data Science and AI - A chat with Computational Linguist Jonathan Manczur
Paterson, NJ - On Feb. 23, 2023, Jonathan Manczur, Professor of Computer Information Systems, was interviewed regarding Artificial Intelligence. DV: What is AI?
JM: I love teaching AI because I love disappointing students with the answer to that. Not that I actually enjoy disappointing students, but the idea is [that] AI is not consciousness: it's not self-thinking robots. What it really is [is] wrapped up in computational statistics, running statistical algorithms to make predictions or to categorize things. Computers are much better at math than we are, [and] we can't pretend to be a computer. Eventually, we can get to the mathematical formula that [a] human will take time to [solve, but] then the computer will do it instantaneously. The question of artificial intelligence is how do we think as humans the way that computers can't think, and there are [a] number of different things that we can do, [that computers cannot]: we can answer emotional questions, we can answer abstract language and riddles, things like that. Can a computer ever get to that point? AI’s intending to replicate that level of humanity, that level of intelligence. Human intelligence.
DV: Yes, so the only thing that really differentiates us with computers they can do math faster, but we can respond to emotional questions.
JM: Yes, there's a movie from the 80s named Blade Runner that had a fantastic thing: to check to see if someone was a robot, they’d ask questions like, "you're walking in a dessert, and you see a turtle on its back, why aren’t you helping it?" [In the film], the computer can't answer those questions. A human can be like, "Alright, I can reason this out." That’s the difficulty, and has been the difficulty, of chat bots: they can't handle complex [or] abstract questions things like that.
DV: Where do you think AI is heading?
JM: Right now, everyone and anyone in business wants to apply AI because that’s the hot topic. So, I think that AI is going to [go] wherever they can fit it in for doing mathematical predictions and statistical probabilities. It's going to have its uses. There are two levels here: the financial aspects of it being like, "Hey, I use AI," and they will get funding, [and] then there's the research aspect, of things like self-driving cars, ChatGPT, or chat AIs.
DV: When you said ChatGPT, Professor Hernandez told me to ask you about it because you knew a lot more about it. He said it's being used by Microsoft in order to get an advantage over Google. Do you think it will?
JM: Maybe. I'm not going to say [anything] definitive because I don’t know the [context]. So, right now Google is being investigated for antitrust [violations], so Microsoft may get the advantage if the government steps in and decides to break up Google. Microsoft underwent an anti-trust lawsuit in the 90s. Everyone is going to attempt to create a ChatGPT clone. Google is also working on quantum computing, so if ChatGPT is [used by] Microsoft, Google might overtake them in different fashion. Jeff Bezos and Amazon, or Elon Musk with Tesla, might [also] come up with their own versions that might take over the market, but I think everyone is right now in a race for the best.
DV: I understand that AI has already been used here at PCCC. How has it been used and what other ways can it be used?
JM: There a field called data science, which is the use of artificial intelligence and statistics to create predictions. We have a data scientist who works for the school. I think there's actually a team of them, and what they do is, using features like gender, race, [socio]economic status, [predict] which students will graduate [and] which won't. [It allows us to ask] how do we reach out those who potentially might not graduate, [or] which areas should we do recruiting from. Those are the kinds of aspects of use AI is [able] to make predictions based on. Here’s the data and lets model it out, and what results do we come up with that. That’s how it's used everywhere, but [then we ask] how do we incorporate beyond that.
DV: How can it be used in English Composition classrooms, in those kinds of classrooms where it's not really mathematical, how can it be used then?
JM: That’s what I studied. It is called computational linguistics, which is the melding of computers and language. There's actually a big fear right now [that with] ChatGPT, [students] can say, "write me a 5 page paper on [a topic]," and [the AI] will come up with it. They are already starting to beat that. I'm actually giving a presentation with faculty about that, but I like to think that it’s a tool. There's [also a] big debate about AI art right now.
DV: I’ve heard about it a lot on Instagram. Artists are concerned because there are a lot of people that can make artwork similar to theirs and sell it without the artist getting paid.
JM: That’s true, but that’s the danger. I was actually talking to an artist about this last weekend: people feared the camera at first. [They said] that['s] not art: you're not drawing, you're not painting a scene, all you're doing is...
DV: Capturing the moment?
JM: Yes. So that was, and still is, considered a debate in art circles: whether photography is an art or not. I think it is, and I think most people would say it is. AI art is simply a new tool and people fear it. What is art, new art, but crafting and mending of past styles. That’s what new AI art would create, very much like I think you can use ChatGPT to write poetry. [You can] be like: here's a style I want [or] here’s a topic I want, and that’s a way of producing. It’s a new tool for producing. I think of like rhyme zone or something like that for doing for people that want to do rhyme schemes for music yea that or maybe now you can go to ChatGPT and go thinking of lyrics for a song there's actually a fascinating one where one people have taken the lyrics of songs put it through the AI art production and they have created music videos doing that
DV: What fascinates you about AI?
JM: Well, I do have one last thing to cover about that. I said it wasn’t a consciousness: it wasn’t about intelligence and the thoughts. I really hope I'm wrong, and that’s what really fascinates me. What happens if we birth a new intelligence? I'm really excited about that, and I studied a lot of philosophy on that and I think that’s very cool.
Further questions regarding Artificial Intelligence can be sent to Jonathan Manczur at [email].