In recent months, a new AI-powered chatbot has made its way to the forefront of conversations about innovative technology. However, ChatGPT is just one example of how artificial intelligence and machine learning seem poised to become as integral to business as the internet and the computer. While many embrace the advances of AI and automation, others still struggle with the looming fear that their jobs may one day be replaced by AI. On this episode of FRB, we take a dive into AI and automation, and how it may influence industries in the near future.

Meet Our Hosts

Erin Camp Associate, San Antonio Twitter: @BusinessLawyerE Follow on LinkedIn »

Art Cavazos Partner, San Antonio Twitter: @FinanceLawyer Follow on LinkedIn »

Episode Transcription

Erin Camp: Hi, I’m Erin Camp, a corporate finance lawyer with Jackson Walker.

Art Cavazos: And I’m Art Cavazos, a corporate and finance lawyer with Jackson Walker. And this is Future Ready Business.

Erin Camp: As always, we’d like to remind our listeners that the opinions expressed today are ours, and do not necessarily reflect the views of Jackson Walker, its clients or any of their respective affiliates.

Art Cavazos: This podcast is for informational and entertainment purposes only and does not constitute legal advice. Now, we’ve got a great show lined up for you today, which is every time and it’s always true. Today, we’re talking about automation and AI. And this is a topic that we’ve wanted to do really, since the outset of this podcast, but recently, kind of late 2022 to early 2023. When we’re recording this, the big thing, the big hotness in AI has been ChatGPT. That’s kind of what everybody’s talking about right now. But we wanted to go really a little more broader in depth than just ChatGPT. And talk about AI and automation in general, I think.

Erin Camp: Yeah, I mean, I think AI is something that kind of people in the technology space have been really trying to achieve specifically artificial general intelligence, which we’ll talk about in a little bit about what that is. But I don’t think many people think about this. But so many of us encounter automation and AI every day, or a subset of that artificial general intelligence every day.

Art Cavazos: Yeah, it’s already all around us.

Erin Camp: Yeah, it’s already all around us. And so I think it’s important first to know what they are, automation is different from AI, automation is sort of like a limited set of possibilities that you program into something and then based on certain inputs, that the machine or the computer that you preprogrammed those limitations to those possibilities to will react to each input a specific way, based on how you programmed it, it just has data in there, and it processes it, it doesn’t try to understand it or do anything additional.

Art Cavazos: Right. And you know, some examples of that we use all the time, think of a dishwasher or you know, anything, you push a button and it just you know, does it for you, you know, even our coffeemaker now, like you can’t make a cup of coffee without, you know, having a computer involved now, and it being you know, automated, but that’s different from artificial intelligence, which, like you were saying is really a very broad term, it can mean a lot of different things. There’s aspects of it that are sometimes called weak AI, meaning it’s really just a subset of artificial general intelligence, which you mentioned, that would be kind of an encompassing kind of umbrella term for a generally intelligent machine, which really, you know, doesn’t exist yet. Right. That’s something that that people are still working towards. We mentioned ChatGPT, their parent company, OpenAI, which has a very interesting history itself, we can go into at some point, but their stated goal is to create artificial general intelligence. So there’s definitely people working on it. That’s a goal of a lot of really big tech companies and tech investors.

Erin Camp: And I think one way to like really explain what Artificial General Intelligence is, is it’s, we’re trying to basically recreate human intelligence. And there’s different subsets different things that we do every day, a different ways we synthesize information, which would be these like different sort of subsets or types of AI or weak AI, that are things.

Art Cavazos: Yeah, things like natural language processing. So just being able to understand the words that I’m saying right now and do something with those, especially if it’s like a question or a command, then, you know, the computer being able to respond to that appropriately, then there’s machine learning, which is the ability to take data, perceive things from it, and intuit the next step, or make a prediction as to what the next input or output should be. So there’s a whole bunch of different subsets of artificial general intelligence, and then you bring that all together. Computer Vision is another one that’s for like self-driving cars, you know, a computer actually kind of being able to see, that’s another type of intelligence. You bring all of that together, and you get Artificial General Intelligence.

Erin Camp: Yeah, and kind of something recently that’s happened to me that just real life really showed me what the difference between automation and AI is. This is so stupid, but I drive a Tesla. And I was one of the idiots, depending on who you ask. That paid for the full AI package because I’m completely fascinated by AI. I mean, that’s why we’ve been wanting to talk about this the whole time. I bought this Tesla and I have AI I had this AI package, but my little sister wanted to try to drive it and we were at a town and had to rent a car. So she wanted to rent one of those like Tesla’s at Hertz. It’s so we rented a Tesla, and my sister was driving it. And it did not have the AI package. But it did have the like hands-free steering and the like smart cruise control.

Art Cavazos: So you were roughing it, but you able to get by.

Erin Camp: Yeah, totally roughing it, it was really hard out there. But you know, we survived. I guess when I think about automation, I think about it as being something that is like not so like, I guess robust is the word I used earlier when I was talking about this with Art, but I just didn’t. But like the hands-free steering and that smart cruise control. That’s not really AI mean, that’s, that’s more automation, for example, with the hands-free steering, and it can stay in the in the lanes. And with the smart cruise control, it can sense the speeds of the cars around you and, and things like that.

Art Cavazos: And most modern cars do have kind of smart cruise control and assist. But I agree totally, it’s not really, it’s definitely not artificial general intelligence. And it’s really not even AI. It’s more like automation.

Erin Camp: Yeah, I would have thought it was kind of AI. But then when I thought about the difference is like it can stay in the lane. And it can react to speeds by slowing down or speeding up depending on how you have it set. But what it can’t do is like for example, change lanes, so you can’t like change lanes in it without the AI package. Because the ability to change lanes. That’s the AI. So the automated part is that smart cruise control in the hands-free steering. But the AI part is like being able to change lanes or exit the highway on time, and things like that. And just so I just think that’s a really good sort of example of what those differences are. That’s the AI part.

Art Cavazos: Yep, totally agree. And good example. So one thing that comes up a lot with AI and automation, I think especially for workers, or maybe white-collar workers, office professionals, because we’ve seen kind of robots kind of take over a lot of manufacturing jobs, for example. And like warehouses become largely automated, but at the same time, like, there’s still a lot of people that work at Amazon. And yet Amazon utilizes robots and automated systems heavily. There is always this kind of interplay between kind of automated and AI resources and work that people are doing.

Erin Camp: And just like are they trying to replace people, I think some people are scared of that, like as a young transactional attorney, and I know like a lot of finance professionals and accountants can relate to this. But the idea of AI started to come out and like legal automation started coming out for like due diligence, for like M&A transactions. And as a younger junior associate, I was like scared for my job, I thought they were like, you know, they’re gonna take my job and replace me with a computer. Now further on in my career, I like would jump at the idea of a computer doing, you know, that due diligence for me and taking on that task. So I mean, you know, depending on what your role is, and what type of role you have with the company, and whether your job is something more automated or not, I think everyone’s kind of scared of AI a little bit when it comes to that.

Art Cavazos: You hit the nail on the head. Look, everyone like literally everyone, because with kind of the ChatGPT discourse that has been going around in early 2023. Part of what I’ve been hearing people talk a lot about is how is this going to impact its use in schools, and the types of concerns you were just talking about as an attorney, which I think are also applicable for a lot of like, finance professionals and all sorts of office work is the exact same concerns that people are talking about at the K through 12 level, or even the college level or really any graduate level. It’s applicable everywhere, where the ability to have the AI kind of do the work for you. Makes it’s very complicated. Because, you know, on the one hand, there’s kind of that fear of, you know, am I going to be replaced? And then on the other hand, I think there’s what you expressed, and I agree a lot with, which is if we can get rid of a lot of these menial tasks, and people just don’t have to do those tasks like everybody benefits, right. And people can focus on more substantive tasks, which I think is usually what people would rather be working on anyway. Very few associates, relish, diligence work, even if they could be spending time on more substantive work.

Erin Camp: Well, I do think like, you know, what’s really interesting about open AI is not only is their stated purpose, and I think for this reason, not only is their stated purpose like that they want to come up with general, what were we calling it artificial general intelligence, but they want to do it, they want to do it in like an ethical way that helps humanity instead of like, replacing humanity, I guess is the idea there. That’s not the words they use, replacing, but I feel like maybe that’s kind of the, obviously the fear,

Art Cavazos: Right, because there is, I think, a path that can be envisioned and can be laid out that this actually is all like very helpful to humanity. And we just all get to do less menial tasks and get to focus on more substantive work. And also, there’s a huge population out there that is underserved in pretty much every way, right? If you want to talk about legal services, if you want to talk about financial services, if you want to talk about just basic needs, like clean food, and water and things like that, there is a huge global population out there that is underserved right now. And so if AI is able to free up workers to work for more people, and kind of provide those services and provide those products to the people that need them, and we have less of an underserved population, that that’s a great use for AI.

Erin Camp: Before we move on, from that thought kind of talk, like bring it back to also sort of like how similar this is to when like personal computers came about, because that is very positive. We’re going to try to help everybody we’re going to do whatever. But whenever computers came out, people were scared. I mean, a lot of people thought their jobs are over. I think, you know, we’ve talked about this before, that happens a lot. There’s these hype cycles, we called it, I think it’s referred to as an Amara’s Law. But we were I think we renamed it and one of the episodes as an Amara’s observation.

Art Cavazos: It’s more like an observation. It’s not really a law.

Erin Camp: Yeah. But it does seem to happen over and over again. I mean, I think a lot of people think of this as like the next step after computers and technology progression. It’s just kind of happening again, we’re really excited about it. We’re in the hype cycle right now. ChatGPT just came out, we’re wondering how it’s going to change our lives for like everybody, I’m really interested to see how like how we go forward, in terms of people’s feelings about AI, specifically, people scared about their jobs. (SNIPPET)

Art Cavazos: And I think that is a good point to point out about ChatGPT being kind of in that Amara’s Law or observation hype cycle right now, which is usually comes at the beginnings, basically, new technology is introduced, there’s this huge hype around it, people think that it’s just going to change the world overnight, nothing will ever be the same. And then of course, the reality kind of sinks in and people kind of swing the other way and kind of become disillusioned with the technology, see all of its flaws. And you know, the drawbacks and things like that, then at some point, it kind of evens out. And the new technology is used, kind of in a more normalized way. And so I do think right now, with ChatGPT, we’re in that hype cycle. And I also don’t think that was by accident, because they also were preparing for a very large fundraise at the beginning of 2023. And which was incredibly successful, and probably in no small part due to all of the online chatter and hype that they were able to garner very successful kind of hype cycle there that I think was probably to some extent orchestrated.

Erin Camp: I am curious, you know, like, besides like speeding up menial tasks and things, like what others sort of arguments there are in favor of AI? Because I feel like when I hear conversations about AI, they’re almost solely around, is it going to take my job? Is there enough work out there for people? If we invent something that can take over all these menial tasks. In a planet where we are very overpopulated. Is that going to take away the ability of some people to work? And do all these things. How do we get around that?

Art Cavazos: It’s definitely a very complicated issue, and there’s probably a million different.

Erin Camp: And clearly, I’m not expecting you to answer that.

Art Cavazos: Yeah, I just want to like caveat that I’m not going to, you know, give the definitive answer. But some arguments in favor of AI that you could articulate, or that are possible. You know, for example, there could be a greater we were talking about more substantive work, you know, and what I mean, really, by substantive, you know, that could be more creative work, things that really take a more human approach. So a lot of times when we think about a job or work, there’s so much involved with it, that really isn’t that much fun, right. It really isn’t that pleasant, but it’s got to get done because it’s part of the job. The more you can automate and hand off to AI, those parts of the job, and then just focus on the like fun parts of the job. And I think for a lot of people, the fun parts of the job involve other people. You know, you brought up diligence. If you were able to have the AI do a lot of the doc review and pulling out the salient points. And then you hop on a call with opposing counsel and go through those and have that discussion. That part is probably a little bit. It depends on the opposing counsel, right.

Erin Camp: But you’re getting rid of all the associates.

Art Cavazos: Well, so that’s the point I did want to bring get up earlier is that so when you think about the K through 12 Arena, I think one of the fears there about a tool like ChatGPT is that if you allowed, let’s say, elementary students to use it, or maybe you didn’t even allow them, but they just went ahead and started using ChatGPT, to help them with their assignments, help them with their homework and that sort of thing. Then they may never even learn kind of the basic skills that you need, in order to write an essay, or in order to do a lot of the basic things that we think of learning in school. It seems like, you need to kind of restrict it in some way where you say, okay, maybe there are places for it to be used. But it’s kind of like not letting you use a calculator and kind of forcing you to take the test by hand and do it pencil and paper, there may be still arenas where that is beneficial, that we actually want the people to go through and kind of do it by hand, and learn how to do it by hand, you know, almost like for training purposes, right.

But once those training purposes are addressed, I think that there is a very strong argument that once you have a calculator, you’d rather use the calculator. For learning and testing purposes, maybe you want to have restrictions on it, but out kind of in the wild, or you know, in the real world where you really just want to get things done the most efficient way possible, then you probably do want to go ahead and utilize the technology utilize the tools, but you probably want to have some of the same considerations around should we be having, you know, Junior associates do some of this as like a learning exercise? Or are there other ways for them to learn, other than doc review, and that sort of thing, that the AI can do faster and more efficiently and cheaper. And so maybe you addressed the training concerns in a different way, without having to force an associate to go through two years of nonstop doc review, as their, you know, first couple years as an attorney.

Erin Camp: Yeah, it’s really interesting to think about too, because it’s like, the final result is this, like, artificial general intelligence that is fundamentally different than automation. That doesn’t depend on this like realm of possibilities. It’s kind of interesting to think about, like how that’s even really possible. Because even now, with AI like that, like, like, let’s go back to the Tesla, or even, I mean.

Art Cavazos: Siri, Alexia, ChatGPT.

Erin Camp: Yeah, they all like the way they function. And their ability to function is big, just by virtue of having so many users. So they have all these users spending all these possibilities. So it’s kind of like, even with ChatGPT, have they really even is that really even still AI? Or is it still depending on a realm of possibilities? Because I think, really, to achieve true AI, you really have to exist outside of limitation. It’s really interesting to think about what that means, you know, the human mind, it just seems like it is without limit, it’d be very interesting to see how, or if we can ever figure out a way to replicate that.

Art Cavazos: Yeah, and maybe not, but you know, as far as the limitations, I mean, I agree, but the human mind, you know, he is capable of a lot. But there’s definitely limits, right? Like, if you put me in a test in a completely foreign language that I have no familiarity with, and all of the instructions are in a foreign language. I wouldn’t even know where to start. But I agree with your point that like, I could probably go and learn that language. Like if it was really imperative, I could spend the time and learn that language, and then I could go and take that test. And so that’s not a full limitation. And I think with an artificial general intelligence, which, which we don’t have yet, but companies like OpenAI are working towards is something very similar where like, yes, it would have certain limitations be stuck within those realm of possibilities, as you were saying, but all that would really need is additional data, you know, essentially, like me going and learning that language, it would just need to access those additional resources that additional data, and a true AGI, artificial general intelligence, would be able to surpass that limitation, once it has access to that additional data, it would synthesize that learn and adapt.

Erin Camp: But I think that’s also like a really good point. And also like, kind of what we’re all scared of is, yes, that is the exercise that AGI would be doing. But if it’s like a computer, it already has unlimited access to unlimited information that a human being could never possess. If you were to really create AGI like the capability of a computer to realize they don’t know something and they need more information in order to make a decision. And then they creatively figure out which data they needed, and they put a decent size it all together and they come up with whatever the answer or solution, whatever they’re doing. I mean, if the limitation for us is our inability to store unlimited information, and then this like computer has access to unlimited information, I think that’s why AI is scary to people. And I think that’s why it can be scary to everyone, like we said earlier to like, no matter what your job is, because I think some people and a lot of people in tech that are working so hard to achieve AGI, I do think they lose sight of that it really could take any job away as we see them today.

We think about things like, recently I heard like truck drivers, that could be a job that becomes extinct if we moved to fully self-driving cars, things like that. I mean, that’s what I think of when I think of a computer taking my job. But it really, I mean, ChatGPT was like passing medical licensing exams. And I read on an article that they also passed the bar exam. Hilariously, that’s not true.

Art Cavazos: Parts of the bar exam.

Erin Camp: Yeah, only parts. I love that Art jumped in there. listening to this podcast would do the same thing. But I want to point out it was the MBE. And I’m sorry. That’s the hardest part of the bar exam. It’s been a while since I’ve taken it. But that is 100% The hardest part. And it passed it?

Art Cavazos: Well, that’s the part that’s multiple choice, isn’t it?

Erin Camp: Yeah. But yeah, but I mean, like ChatGPT can write like essays.

Art Cavazos: No, it can.

Erin Camp: That’s actually like my best friend for her brother’s wedding. Used ChatGPT to write her bridesmaid speech. And I think that’s hilarious. And it was accurate.

Art Cavazos: So something occurred to me the other day, because we make this podcast, there’s a lot of podcasts out there. There’s a lot of websites out there, there’s a lot of social media out there, you know, people have kind of put all of this content, essentially onto the internet. And ChatGPT uses this model that has basically scraped the internet and pulls all sorts of things from photos to you know, text. So I mean, I definitely envision a world where you could type into ChatGPT, create a podcast episode in the style of Art and Erin on the topic of AI automation. I don’t know how that makes me feel and like, you know, but it would be kind of interesting to see what it would spit out. But you know, the more kind of content evoke types that we put online, that all becomes kind of free rein for these artificial intelligence models to kind of scrape that data and add it to their model.

Erin Camp: But we talk about a lot how like, it seems like people that are in the race to be the most powerful, like, what their tactic is, is to get the most data that they can.

Art Cavazos: Right, because data is very valuable. So I think a lot of companies are thinking about ways to gather more data. And it did also strike me as a very interesting with ChatGPT, that when people are typing in their commands and queries to ChatGPT, I don’t know if they’re always thinking about what is being done with those inputs that they’re giving to ChatGPT.

Erin Camp: Like, when my friend did the interview, she probably said some pretty personal stuff to come out with that speech.

Art Cavazos: Right. Right. And it may not be harmful or insidious things like it reminds me of a friend who texted me and another friend of mine, we’re all three friends in college. And you know, he sent us a text that had been created by ChatGPT. And I was like, just out of curiosity, like what did you tell it? And you know, he basically told that, you know, all this stuff about three of us that, again, it wasn’t like insidious stuff. But it was just stuff that nobody else would know. Other than three of us, and that had not previously been posted to like social media or online, basically. So it was like, it’s just like another avenue by which data is being added to the database through all these inputs and queries.

Erin Camp: Ominous.

Art Cavazos: Well, I think we’ve talked a lot about automation and AI today, and ChatGPT and other aspects of AI and automation. And your Tesla.

Erin Camp: I gotta slip it in there, you know.

Art Cavazos: Alright, so thank you, everyone for joining us on this episode of Future Ready Business. We touched on a lot of things today regarding AI, automation and subsets of it. Hopefully, we’ll be talking about this topic again more soon. We wanted to talk about other angles on it. For example, legal automation is an area that obviously we’re very interested in and have friends and colleagues who would be really great to come on as guests. So be on the lookout for that. And if you’d like to show rate and review us wherever you listen to your favorite podcasts and share Future Ready Business with your friends and colleagues.

Erin Camp: Thank you to our producer Greg on ones and twos. You can find me Erin Camp on Twitter at @BusinessLawyerE. E for Erin.

Art Cavazos: And you can find me on Twitter at @FinanceLawyer. As mentioned at the top of the show the opinions expressed today are ours and those of our guests, if any, and do not necessarily reflect the views of Jackson Walker, it’s clients or any of their respective affiliates.

Erin Camp: This podcast is for informational and entertainment purposes only and does not constitute legal advice. We hope you enjoyed it. Thanks for listening.

Art Cavazos: Goodbye, everybody.

Visit JW.com/future-ready-business-podcast for more episodes. Follow Jackson Walker LLP on LinkedIn, Twitter, Facebook, and Instagram.

This podcast is made available by Jackson Walker for informational purposes only, does not constitute legal advice, and is not a substitute for legal advice from qualified counsel. Your use of this podcast does not create an attorney-client relationship between you and Jackson Walker. The facts and results of each case will vary, and no particular result can be guaranteed.