Unveiling The Future Of AI

0:00Hey, did you know that HubSpot's annual Inbound conference is coming up? That's right. It'll be in Boston from September 5th through the 8th. Every year, Inbound brings together leaders across business, sales, marketing, customer success, operations, and more. You'll be able to discover all the latest must-know trends and tactics that you can actually put into place to scale your business in a sustainable way. You can learn from industry experts and be inspired by incredible spotlight talent. This year, the likes of Reese Witherspoon, Derek Jeter, Guy Raz are all going to make appearances. Visit inbound.com and get your ticket today.

0:43You won't be sorry. This programming is guaranteed to inspire and recharge. That's right. Go to inbound.com to get your ticket today.

1:02Hello, and welcome to another episode of the Duct Tape Marketing Podcast. This is John Jantz. My guest today is Kenneth Wenger. He's an author, research scholar at Toronto Metropolitan University and CTO of Squint AI Inc. His research interests lie in the intersection of humans and machines, ensuring that we build a future based on the responsible use of technology. We're going to talk about his book today is The Algorithm Plotting Against Us, A Layperson's Guide to the Concepts, Math, and Pitfalls of AI. So Ken, welcome to the show. Hi, John. Thank you very much. Thank you for having me. So we are going to talk about the book, but I'm just curious, what does Squint AI do?

1:47That's a great question. So Squint AI is a company that we created to do some research and develop a platform that enables us to do AI in a more responsible way. okay so uh i'm sure we're going to get into this but i touch upon it uh in the book in many cases as well where we talk about uh ai ethical use of ai some of the downfalls of ai and so what we're doing with squint is we're trying to figure out you know how do we try to create a an environment that enables us to use ai in a way that lets us understand when these algorithms are not performing at their best when they're making mistakes and so on.

2:30Yeah. So, so the title of your book is the algorithm plotting against this is a bit of a provocative question. I mean, obviously I'm sure there are people out there that are saying no, and some are saying, well, absolutely. So, so why ask the question then? Well, because I actually feel like that's a question that's being asked by many different people with actually with different meaning, right? So So it's almost the same as the question of is AI posing an existential threat? It's a question that means different things to different people. So I wanted to get into that in the book and try to do two things.

3:08First, offer people the tools to be able to understand that question for themselves. And first, figure out where they stand in that debate. And then second, also provide my opinion along the way. Yeah. And I probably didn't ask that question as elegantly as I'd like to. I actually think it's great that you asked the question, because ultimately what we're trying to do is let people come to their own decisions rather than saying this is true of AI or this is not true of AI, right? That's right. That's right. And again, especially because it's a nuanced problem and it means different things to different people.

3:44So this is a really hard question, but I'm going to ask you, where are we really in the continuum of AI? I mean, people who have been on this topic for many years realize it's been built into many things that we use every day and take for granted. Obviously, ChatGPT brought on a whole nother spectrum of people that now, you know, at least have a talking vocabulary of what it is. But I remember, you know, I've been I've been I've had my own business 30 years. I mean, we didn't have the Web. We didn't have Web sites. You know, we didn't have mobile devices that certainly now play a part. But I remember as each of those came along, people were like, oh, we're doomed.

4:20It's over. Right. So currently there's a lot of that type of language surrounding AI. But where do you think we really are in the continuum of the evolution? You know, that's a great question, because I think we're actually very early on. I think that, you know, we've made remarkable progress in a very short period of time. But I think it's where the very early stages, you know, if you think of AI where we are right now and we were a decade ago we've made some progress but i think the fundamentally at a scientific level we've only started to scratch the surface i'll give you some examples so initially you know the first models that were great at really giving us some proof that this new way of posing questions you know the neural networks essentially yeah right they're very complex equations.

5:09If you use GPUs to run these complex equations, then we can actually solve pretty complex problems. That's something we realized around 2012. And then after around 2017, so between 2012 and 2017, progress was very linear. New models were created, the new ideas were proposed, but things scaled and progressed very linearly. But after 2017, with the introduction of the model that's called the transformer, which is the base architecture behind ChatGPT and all these large language models, we had another kind of realization. That's when we realized that if you take those models and you scale them up, and you scale them up in terms of the size of the model and the size of the data set that we used to train them, they get exponentially better.

5:57Okay, and that's when we got to the point where we are today, where we realized that just by scaling them, again, we haven't done anything fundamentally different since 2017. team, all we've done is increase the size of the model, increase the size of the data set, and they're getting exponentially better. So multiplication rather than addition. Well, yes, exactly. Yeah. So it isn't, the progress has been exponential, not only in linear trajectory, but I think, but again, the fact that we haven't changed much fundamentally in these models, that's going to taper off very soon. It's my expectation.

6:31And now, where are we on the timeline, which was the original question? I think if you think about what the models are doing today, they're doing very simple statistics, essentially. The idea of these models being called artificial intelligence, I think it's a bit of a misnomer sometimes. I agree. And it leads to some of the questions that people have because there isn't much deep intelligence going on. It's just statistical modeling and very simple at that. And then where we're going from here and what I hope the future is, that's when we start. I think things are going to change dramatically when we start getting models that are able not just to do simple statistics, but are able to understand the context of what it is they're trying to achieve.

7:21And are able to understand the right answer as well as the wrong answer. So, for example, they're able to know when they're talking about things they know and when they're kind of skirting around this gray area of things they don't really know about. Does that make sense? Yeah, absolutely. I mean, I totally agree with you on artificial intelligence. I've actually been calling it IA. I think it's more of informed automation is kind of how I look at it, at least in my work. Do you see a day where prompts, asking questions, or that's kind of the street use, if you will, of AI for a lot of people.

7:57Do you see a day where it starts asking you questions back? Like, why would you want to know that? Or what are you trying to achieve by asking this question? Yeah. So the simple answer is yes, I definitely do. And I think that's part of what achieving a higher level intelligence would be like. It's when they're not just doing your bidding. It's not just a tool, but they kind of have their own purpose that they're trying to achieve. And so that's when you would see things like questions essentially arise from the system, right? It's when they have a goal they want to get at, and then they figure out a plan to get to that goal.

8:34That's when you can see emergence of things like questions to you. I don't think we're there yet, but I think it's certainly possible. But that's the sci-fi version too, right? I mean, where people start saying, you know, the movies, it was like, no, no, Ken, you don't get to know that information yet. I'll decide when you can know that. Well, you're right. I mean, the question, the way you asked the question was more like, is it possible in principle? I think absolutely yes. Do we want that? I mean, I don't know. I guess that's part of, it depends on what use case we're thinking about. But from a first principles perspective, yeah, it is certainly possible to get a model to do that.

9:14So I do think there are scores and scores of people. Their only understanding of AI is I go to this place where it has a box and I type in a question and it spits out an answer. Since you have both layperson and math in the title, could you give us sort of the layperson's version of how it does that? Yeah, absolutely. So, well, at least I'll try. Let me put it that way. A few moments ago, when I mentioned that these models, essentially what they are, they're very simple statistical models. That phrase itself, it's controversial because at the end of the day, we don't know what kind of intelligence we have.

9:53So if you think about our intelligence, we don't know whether at some level we are also a statistical model. right however what i mean by ai today in large language models like chat gpt being simple statistical models what i mean by that is that they're performing a very simple task so if you think of chat gpt what they're doing is they are trying essentially to predict the next best word in a sequence that's all they're doing and the way the way they're doing that is that they calculate what are called probability distribution. So basically for any word in a prompt or in a corpus of text, they calculate the probability that word belongs in that sequence, right?

10:41And then they choose the next word with the highest probability of being correct there. Okay. Now that is a very simple model in the following sense. If you think about how we communicate, right? You know, we're having a conversation right now. I think when you ask me a question, I pause and I think about what I'm about to say, right? So I have a model of the world and I have a purpose in that conversation. I come up with the idea of what I want to respond. And then I use my ability to produce words and to sound them out to communicate that with you, right? It might be possible that I have a system in my brain that works very similar to a large language model in the sense that as soon as I start saying words, the next word that I'm about to say is one that is most likely to be correct given the words that I just said.

11:32It's very possible that's true. However, what's different is that at least I already have a plan of what I'm about to say in some latent space. I've already encoded in some form what I want to get across. how I say it that the ability to produce those words might be very similar to a large language model but the difference is that a large language model is trying to figure out what it's going to say as well as coming up with those words at the same time right does that make sense so it's a bit like they're rambling and sometimes if they talk for too long they ramble in a nonsense territory because they don't know what they're going to say until they say it.

12:17So that's a very fundamental difference. Yeah, I have certainly seen some output that is pretty interesting along those lines. But, you know, as I heard you talk about that, I mean, in a lot of ways, that's what we're doing is we're querying a database of what we've been taught. The words that we know, in addition to the concepts that we've studied and are able to articulate, I mean, in some ways, we're querying that to me prompting or me asking you a question as well i mean it works similar would you say the aspect of prompting a question and then answering it it's similar but what is different is the the concept that you're trying to describe so again when you ask me a question i think about it and i come up with so again i have a world model that works so far for me to get me through life And that world model lets me understand different concepts in different ways.

13:13And when I'm about to answer your question, I think about it. I formulate a response. And then I figure out a way to communicate that with you. That step is missing from what these language models are doing. They're getting a prompt. But there is no step in which they are formulating a response with some goal. right? Some purpose. They are essentially getting a text and they're trying to generate a sequence of words that are being figured out as they're being produced, right? There's no alternate plan. So that's a very fundamental difference. And now let's hear a word from our sponsor, Marketing Made Simple.

13:58It's a podcast hosted by Dr. JJ Peterson and is brought to you by the HubSpot Podcast Network, the audio destination for business professionals. Marketing Made Simple brings you practical tips to make your marketing easy and more importantly, make it work. In a recent episode, JJ and April chat with StoryBrand certified guides and agency owners about how to use ChatGPT for marketing purposes. We all know how important that is today. Listen to Marketing Made Simple wherever you get your podcasts. Hey, marketing agency owners, you know, I can teach you the keys to doubling your business in just 90 days or your money back.

14:38Sound interesting? All you have to do is license our three-step process that's going to allow you to make your competitors irrelevant, charge a premium for your services, and scale, perhaps without adding overhead. And here's the best part. You can license this entire system for your agency by simply participating in an upcoming agency certification intensive. Look, why create the wheel? Use a set of tools that took us over 20 years to create, and you can have them today. Check it out at dtm.world slash certification. That's dtm.world slash certification. I do want to come to like what the future holds, but I want to dwell on a couple things that you dive into in the book.

15:25What are the, you know, other than sort of the fear that the media spreads, what are the real, you know, and obvious pitfalls of relying on AI?

15:38I think the biggest issue and one of the, I mean, the real motivator for me when I started write in the book is that it is a powerful tool for two reasons. It's very easy to use seemingly, right? You can spend a weekend learning Python. You can write a few lines and you can transform, you can analyze, you can parse data that you couldn't before just by using a library. So you don't really have to understand what you're doing and you can get some result that looks useful. Okay. But hidden in that process, right? The fact that you can take data, a lot, a large amounts of data, modify it in some way and get a response, get some result without understanding what's happening in the middle has huge repercussions for misunderstanding the results that you're getting.

16:35Right. And then if you're using these tools in the world, in a way that can affect other people, for example, let's say you work in a financial institution and you come up with a model to figure out who you should give some credit, a proof for credit, for a credit line, and who you shouldn't. Now, right now, banks have their own models. But if you take the AI out of it, traditionally, those models are thought through by statisticians. And they may get things wrong once in a while. But at least they have a big picture of what it means to analyze data, biasing the data. What are the repercussions of biasing the data?

17:21How do you get rid of all these things are things that a good statistician should be trained to do. But now if you remove the statisticians, because anybody can use a model to analyze data and get some prediction, then what happens is you end up denying and approving credit lines for people with, you know, with repercussions that could be, you know, driven by very negative bias in the data, right? Like it could affect a certain section of the population negatively. Maybe there's some people that can't get a credit line anymore just because they live in a particular neighborhood or they, you know, there's many reasons why this could be a problem.

17:57But wasn't that a factor previously? I mean, certainly neighborhoods are considered, you know, as part of the, you know, even in the analog models, I think. Yeah, absolutely. So, like I said, we always had a problem with bias, right, in the data. But traditionally, you would hope, so two things would happen. First, you would hope that whoever comes up with a model, just because it's a complex problem, they have to have some statistical training. And an ethical statistician would have to consider how to deal with the bias in the data. So that's number one. Number two, the problem that we have right now is that, first of all, you don't need to have that statistician.

18:36You can just use a model without understanding what's happening. Right. And then what's worse is that with these models, we can't actually understand how the or it's very difficult traditionally to understand how the model arrived at a prediction. So if you get denied either a credit line or as I talk about in the book, bail, for example, in a court case, it's very difficult to argue, well, why me? Why was I denied this thing? And then if you go through the process of auditing it, again, with the traditional approach where you have a decision, you can always ask, so how did you model this? Why was this person denied this particular case in an audit?

19:16With a neural network, for example, that becomes a lot more complicated. So I mean, so what you're saying, one of the initial problems is that people are relying on the output, the data. I mean, even, you know, I use it in a very simple way. I run a marketing company and we use it a lot of times to give us copy ideas, give us headline ideas, you know, for things. So I don't really feel like there's any real danger in there other than maybe sounding like everybody else in your copy. But you're saying that, you know, as people start relying on these to make decisions that are supposed to be informed, a lot of times predictions are wrong.

19:57yes and there's very so the answer is yes now there's two reasons for that and by the way let me just go back to say that there are use cases where of course you have to think about this as a spectrum right like yeah there are cases where the repercussions of getting something wrong is worse than other cases right so as you say if you're trying to generate some copy and you know if it's nonsensical then you just go ahead and change it and at the end of the day you're probably going to review it anyway. So that is a lower, probably a lower cost, the cost of a mistake that would be lower than in the case of using a model in a judicial process, for example.

20:32Right, right, right. Now, with respect to the fact that these models sometimes make mistakes, the reason for that is that the way these models actually work is that they, and the part that can be deceiving, is that they tend to work really well for areas in the data that they understand really well. So if you think of a data set, right? So they're trained using a data set. For most of the data in that data set, they're going to be able to model it really well. And so that's why you get models that perform, let's say, 90 % accurate on a particular data set. The problem is that for the 10 % where they're not able to model really well, the mistakes there are remarkable.

21:17and in a way that a human would not be able to make those mistakes. So what happens in those cases is that, first of all, when we're training these models that we get, we say, well, you know, we get 10 % error rate in this particular data set. The one issue is that when you take that into production, you don't know that the incidence rate of those errors are going to be the same in the real world, right? You may end up being in a situation where you get those data points that lead to errors at a much higher rate than you did in your data set. Just one problem. The second problem is that if you're in a, if your use case, if your production, you know, application, it's such where a mistake could be caused, like let's say in a medical use case or in self-driving.

22:03When you have to go back and explain why you got something wrong, why the model got something wrong, and it is just so bizarrely different from what a human would get wrong, that's one of the fundamental reasons why we don't have these systems being deployed across safety critical domains today and by the way that's one of the fundamental reasons why we created splint is to tackle specifically those problems is to figure out how can we create a set of models or a system that's able to understand specifically when models are getting things right and when they're getting things wrong at runtime because i really think it's one of the fundamental reasons why we haven't advanced as much as we should have at this point.

22:44It's because when models work really well, when they're able to model the data well, then they work great. But for the cases where they can't model that section of the data, the mistakes are just unbelievable, right? It's things that humans would never make those kinds of mistakes. Yeah. And obviously, that's certainly going to, that has to be solved before anybody's going to trust sending, you know, a manned spacecraft, you know, guided by AI or something, right? I mean, when, you know, human life is at risk, you know, you've got to have trust. And so if you can't trust that decision-making, that's certainly going to keep people from employing the technology, I suppose.

23:24Right. Or using them, for example, to help in, as I was saying, in medical domains, for example, cancer diagnosis, right? If you want a model to be able to detect certain types of cancer given, let's say, biopsy scans. You want to be able to trust the model. Now, anything, any model, essentially, you know, it's going to make mistakes. Nothing is ever perfect. But you want two things to happen. First, you want to be able to minimize the types of mistakes that the model can make. You need to have some indication that the quality of the prediction of the model isn't great. You don't have that. And second, once a mistake happens, you have to be able to defend that the reason the mistake happened is because the quality of the data was such that, you know, even a human couldn't do better.

24:08We can't have models make mistakes that a human doctor would look at and say, well, this is clearly incorrect. Yeah, yeah, absolutely. Well, Kenneth, I want to thank you for taking a moment to stop by the Duct Tape Marketing Podcast. You want to tell people where they can find, connect with you if you'd like, and then obviously where they can pick up a copy of, is the algorithm plotting against us? Absolutely. Thank you very much, first of all, for having me. It was a great conversation. So yeah, you can reach me on LinkedIn and for a copy of the book, you can get it both from Amazon as well as from our publisher website.

24:40It's called workingfires.org. Awesome. Well, again, thanks for stopping by. Great conversation. Hopefully, maybe we'll run into you one of these days out there on the road. Thank you. Hey, and one final thing before you go. You know how I talk about marketing strategy, strategy before tactics. Well, sometimes it can be hard to understand where you stand in that, what needs to be done with regard to creating a marketing strategy. So we created a free tool for you. It's called the Marketing Strategy Assessment. You can find it at marketingassessment.co, not.com,.co. Check out our free marketing assessment and learn where you are with your strategy today.

25:20That's just marketingassessment.co. I'd love to chat with you about the results that you get.