Harness AI Principles Beyond Generative AI

0:00Hello and welcome to the AI for executives and leaders course my name is Jonathan Barrios, and I'm excited you're here We are living in such an exciting time Artificial intelligence is getting better and better and right now Everyone especially leaders of industry are trying to make sense of how to best use it How to scale it how to leverage it and how to make sure their teams can work most successfully Alongside it so what is this course about what's inside the scope of what we'll be learning and what is outside of it Here's what's included in the course choosing AI wisely. We'll go over how to pick the best AI tools for a Particular business task and then we'll explore automation Decisions here we'll talk in-depth about when to fully automate versus when to use a human in the loop or ITL approach finally we have scaling safely and that's the key word safely We'll start with small real-world AI pilot project examples that demonstrate security Business value and how you can scale them easily. All right, so what's not in the course What's not included and on this side? We have no coding or model building

1:17So you won't learn how to train AI models or write any code This is a non-technical course for executives and leaders and that means no AI tool tutorials No step-by-step guides on AI tools like chat GPT Gemini Claude, etc We will talk about them, but we're not going to dive deep into them and finally we have no deployments We're not going over technical setup and the focus of the course will be on high-level concepts and decision-making And that means we're going to stick pretty close to the concepts that come with implementing AI and before we get into the skill I did want to show you that this is the first course in a series of mini AI courses This one is artificial intelligence for executive and leaders the one we're now However, we do have really interesting courses that are very powerful that can work together for example prompt engineering We'll talk about different types of AI and how prompt engineering is really a form of programming That means when you're talking to chat GPT, you're literally programming in plain language like English So knowing the different techniques of AI prompt engineering will really make

2:32your models much more effective and you can mitigate Problems such as hallucination, which we'll talk about in depth in this course There's also agent decoding and this is really programming with an AI agent And if you use the Microsoft ecosystem Microsoft co-pilot is embedded into Excel word and all of the different aspects of Microsoft's ecosystem Vibe coding is programming for non-programmers leveraging AI and while it can be done Taking this course is really important because there are a lot of dangers that can come up and avoiding those are key And then here are two very important ones AI compliance Privacy and copyright this one over here is more for HR legal and IT So non-technical audiences even though IT is a technical audience This course stays at a high level and really just talked about compliance copyright and privacy while this one over here focuses on language and techniques specifically for developers and finally if you want to get your teams up to speed On AI productivity. This is a great course to get them up to speed and a little bit about me I teach data science machine learning and deep learning and I've worked with

3:50many AI tools and built SaaS websites AI applications and streaming platforms I've been teaching for a while now from treehouse to thinkful and check and now I'm happy to be here with the CBT nuggets team you could also find me here on LinkedIn or my website where post blogs and Social media at AI underscore data underscore science and here at X even though some people don't like it It's really where all of the AI community is in social media So I use it to stay up-to-date with all of the new developments and I ignore pretty much everything else And now we're ready for the first skill in our course where we're going to harness AI principles beyond generative AI and don't worry. We're going to explain what generative AI is and just a quick hint It's more than one thing. We'll start with the fundamental principles of AI itself These are the core technical principles that never change and are always action able even as trends come and go These core principles will empower you with the right tools to integrate AI into everyday processes and workflows In fact, you might be surprised to learn that AI Fundamentally is software modeled after the human brain. That's right. Hey, I

5:09was modeled after us So as we progress through this course, not only will we cover the core principles But we'll go over the different types of AI AI's for core models and how to evaluate which I approach we should be leveraging So as we progress through this course, not only will we cover these core principles But we'll go over all the types of AI AI's for core modes and how to evaluate which AI approach should be leveraged to drive business goals Before we get started, let's talk about everything We're going to cover in this first skill and this first part is the AI toolbox and the takeaway here is that it's not one tool It's many tools inside of this AI toolbox Generative AI is only one option among several and when we say generative AI We're really talking about all of these language models like chat GPT Claude Gemini and grok which uses real-time data then we'll segue into the four modes of AI because again AI is not just one thing here We're going to talk about rules classic machine learning deep learning gen AI

6:25all the good stuff and which model you should use for Which type of task, but how do you do that? And that's why we will talk about decision and safety here. We're going to use a traffic light metaphor of green Yellow and red these are going to be rules that are going to help us Understand if we need a human in a loop for anything that is sensitive or if we can just rely on AI out of the box And finally we have the challenge which is something that we do for all of the skills And what does that mean?

6:56Well, we want to make sure that we understand all of these concepts before we move on to the next skill because they all build on each other progressively Alright, so that's it for this introduction and I'll see you in the first video where we answer the question. What is AI? See you there

0:00Welcome back. In this video, we're going to kick it off by asking the question, "What is artificial intelligence?" But before we answer that question, let me ask you this question. When you hear artificial intelligence or AI, what do you picture? Do you see a chat bot generating big chunks of text? If so, that's normal. Most people do. That's because popular tools like chat GPT make it easy to assume that AI is all generative AI. So the tool generative AI is something that we're going to define. And that's part of models like chat GPT, which makes it easy to assume that all AI is generative AI. So if we ask the question, "What is AI?" we could look at this formal definition here. And it says that it's a field of computer science that focuses on building systems capable of performing tasks that typically require human cognitive abilities. And then we can talk about each one of these. We have perception, reasoning, learning, problem solving, and decision making.

1:05But today, I want to zoom out and away from generative AI, because AI isn't just one thing. It's actually that toolbox we were talking about earlier. And if we understand each one of the tools, we can pick the simplest, most effective tool to improve outcomes for any business. So instead of thinking about human cognitive abilities, we can think about the tools and better yet, outcomes. One of the outcomes we'll explore is onboarding. And we're not really thinking about building an AI tool for onboarding for our organization, but it's a great example. When I onboarded not just to one company, but many companies, there was a recurring theme. If I had a question, I would reach out to somebody. They wouldn't know. They'd have to reach out to somebody else, then give me an email reply, or I'd get a calendar invite and jump on a quick call.

1:59All of this could have been handled very easily with AI, but the trick is AI can get it wrong sometimes. So what do you do? Do you add a human in the loop? Or do you use a specialized AI model that your team will build to make sure that it doesn't say the wrong thing? These are all great questions. And we're going to dive into that in this skill. Okay, so here is generative AI. I promise we'll define this. But for now, let's look at the overall picture. So here we have our AI toolbox. And there's a lot inside of it, right? So gen AI is in the box, but it 's not the whole box.

2:36So the general umbrella term is AI or artificial intelligence. So that's the first one. And then we have machine learning. And then there's deep learning, which you could think of it as a newer type of machine learning. And then inside of that, you have generative AI. And the takeaway here is that AI isn't just generative AI. This is what everyone sees, because these are things like chat GPT. But when is the last time that you saw deep learning or machine learning in action?

3:07Well, those are going to be abstracted. And you're going to see them inside the tools like chat GPT or Gemini. But they're all subsets of each other. So they're interconnected. So here's a quick disclaimer. It's not particularly important to know how all of these types of AI work deeply at a technical level. The takeaway here is more about understanding the tools and how we can use them for different business use cases. So let's take a look at some different use cases that you might see in the real world, and then we'll apply them. So let's take the first one here, facial recognition. This uses deep learning. And you can think of deep learning as a part of AI that is especially good at recognizing patterns. So instead of writing code, I 'm going to use this hamburger and a hot dog here. And we're going to build our own deep learning model. And we're going to use not for facial recognition, but the next one computer vision. And for computer vision, this also uses deep learning. So jumping back to facial recognition, this works because deep learning can spot unique patterns of a person's face, like the spacing of the eyes or the shape of a nose across thousands of photos. And when we're thinking about computer

4:25vision, this is a larger umbrella. So facial recognition fits into computer vision. So it's easy to get mixed up. But here, we're thinking about self driving cars that can see road signs or medical AI. And that's why they both use deep learning, because as I said, deep learning is great at learning patterns. And that allows them to see a make sense of images and details. And the way I like to explain computer vision is that we're teaching the computer to see. We also teach it to read and write. And again, this is how software is modeled after the human brain. When we look at robotics, you could think of this as ML plus DL. So machine learning and deep learning.

5:10So why does it use machine learning robots use machine learning to learn from experience? So not patterns this time. For example, it gets better at gripping objects, walking on uneven ground and optimizing routes in a warehouse. And it uses deep learning when it needs to see with the camera or sensors. So here, it's maybe looking at this cup of coffee and the person that is going to deliver that coffee and that their hand is right here, it's calculating the distance. So robotics is one of the few areas that truly spans both machine learning and deep learning, since it needs learning and perception.

5:47Okay, so now let's look at documents. This is IDP intelligent document processing. And here, this is about teaching the computer to read forms or invoices may be contracts. And here, we're using machine learning, because it learns things like the total amount and how that total amount is usually at the bottom right. But you can also add deep learning, depending on different use cases, for example, deep learning can help when documents are scanned. And then you convert that to clean text. And finally, here we have fraud detection. This is about spotting unusual patterns in data, like let's say a sudden $10,000 charge in another country. ML is right there, because it learns what normal looks like from past transaction. And it flags what doesn't fit.

6:37It learns from historical data. So it learns from the data. So you could think of this as a fraud analyst who's seen millions of transactions and gets a gut feeling when something looks off. But we do this using machine learning instead. And takeaway here is that I haven't mentioned any use case here using generative AI. We could, but this is all deep learning and machine learning, just to show you that a lot of the different things that we interact with that use AI are not generative learning generative learning is one of the tools, such as machine learning and deep learning. Alright, so that's a good stopping point. And I'll see you in the next video on building a mental model. See you there.

0:00Welcome back. Remember when I said that we really didn't need to understand all of these different AI tools at a deep technical level? Well, here's why. My goal with this course is to give you a mental model for applying AI responsibly in business, meaning you can go to a meeting and you can use this information on the fly and name which of the four AI types is in play, connected to the framework and decide where human review needs to stay in the loop. So here is our toolbox and here we can think of this as compliance and here we have our model and here's a human in the loop and then here are the decisions frameworks that we're going to build using the red, yellow and green. And then here's the mental model so that you can use this on the fly and the takeaway of all of this is so that you don't ship poor AI products that are unsecured solutions or worst of all, leak sensitive company data.

1:00Here's how. Instead of thinking about all the different kinds of AI tools, which you will be able to explain after you become familiar with them, I think it's much easier to talk about how software has changed since AI showed up. So we have traditional software, which is software 1.0, you type something into a computer and it's not going to reason. It's not going to use AI. You can say, if this do that and it's always going to do that, it's never going to interpret or think, oh, there's another way I can answer this question.

1:36So that's the first part of the software. The second part is when AI showed up, it started to do things that are using human cognitive skills because it's software modeled after the brain and that's software 2.0. And then finally, we have generative AI, which is software 3.0. So it's much easier to think about AI in the context of the development of software and it's only three different kinds. So we have normal software that just follows instructions that are rules, then we have software modeled after the brain. And number three is really easy to remember because we use plain language. When you talk to chat GPT and you type in a prompt, that's software 3.0. You're literally programming without knowing it. So let's break it down. Software 1.0 is the rules that you write.

2:29And this is perfect when the outcome is clear, if X, then why this could be used for thresholds, lookups or dashboards. A good example is if the ticket is older than 24 hours, escalate because you don't want support tickets to be waiting for a long time. You want to get back to your customers. Now let's take a look at software 2.0. When it first came out, it was mostly machine learning. And machine learning is really about learning from the data. And these are things like algorithms that we can use to, for example, given past tickets, predict the priority one through five. So this is definitely software 2.0. But instead of thinking of it as machine learning, we just say that, well, it's learning from the data.

3:16And the same thing for software 2.0 here. So the difference is that this is early 2.0. And this is really the full version of 2.0, which is deep learning. And deep learning is where we go from algorithms to software models after the brain. For example, read the email body and detect the intent. So right now, if you think about rules and analytics, that's software 1 .0. But if it starts to learn from the data in some kind of way, either through algorithms or by using software modeled after the brain, that's 2.0. Now we have software 3.0. And this is generative AI.

3:55So here we can draft language code and images from instructions. You steer it with prompts rather than changing the code. What you're doing is programming in plain language, for us that is English. An example of generative AI is summarize this thread for the next agent, or maybe draft a welcome email. So that's when we get into generative AI. And what people don't realize is that it's software written in English. And that's a very powerful statement. When you're typing into chat GPT, it's a form of programming. You're doing all of the programming things that you would do in software 2.0, but it's abstracted. It's not as direct, and you can't get as custom as you would in software 2.0. But the output that you get can be very, very similar to 2.0. So that's the power of software 3.0. And that's why prompt engineering is so important, because there are all these little details that go around software 3.0 that are paramount for security, for getting better prompts, for mitigating hallucination. That is where the model will just make things up. So if

5:11you're talking to GPT, sometimes it says things that are not correct, but it's very, very confident. So software 3.0 and prompt engineering go hand in hand. We'll talk more about that as we progress through the course. All right, so let's do a quick review. Software 1.0 are rules in analytics. It's always going to do if X, then Y. For example, if the ticket is older than 24 hours escalated, it's always going to do that. The only reason that there are errors in rule programming, or software of 1.0, which is traditional programming, is when the rules change, and someone updates something, something changes. So that's why you get these errors. But if there 's no errors, or nothing changes, they're going to do this literally forever. Now we have software 2.0, which is both classic ML and deep learning. So let's look at machine learning first.

6:09Here it learns from small patterns from structured data. So scores or labels. In this case, it was past ticket history. And when we're talking about deep learning, it's learning complex patterns from structured and unstructured data. All that means is things like text, images, and audio. It can read the email body and detect intent, but it can also look at an image and say that's a sunrise. And in the upcoming videos, we're going to explore each one of these.

6:39I'm going to show you some code. If X, then Y, so that you can see how it works. And that's pretty straightforward software 1.0. And then we're going to use the hot dog and the hamburger or computer vision. That's software 2.0. And then software 3.0, for that example, we're going to go to chat GPT and produce a similar output. We're going to ask it, what is this image and what is this image. And you're going to see that it's the same output that we got from software 2.0 , where we use a lot of complicated software. But in 3.0, we're using English. And that's to take away. So let's finish with generative AI. And here we simply prompt the model and it just does something. So this could be text code or images. We could say summarize this thread for the next agent or draft a welcome email.

7:31So all of this is done in plain language, in our case, English. And in the next videos, we're going to go over each one of these. And I'm going to show you examples. So this is very, very clear. And this is the foundation that's not going to change. You'll be able to use this no matter what new model comes along. See you there.

0:00Welcome back. In this video we're going to explore software 1.0 and then in subsequent videos we'll go through software 2.0 and software 3.0. So this one is rules and analytics based. Let's take a look and let's consider a very simple scenario when ML or machine learning is not the right solution. So let's say and again this is just a classic probability problem and we have a six-sided die that is numbered one to six and we want to know what the probability of rolling a four is. So we would not use machine learning. The answer is the outcome is deterministic. So we can easily find out what that is which is sixteen point seven percent. So that's not a machine learning problem that is a software 1.0 problem. Meaning regular software. The software we've been using up to today before AI showed up and here's why. Machine learning is not needed because the problem is well-defined and computable. Here is exactly what we're going to do and I'm going to show you this and a couple other examples in Google Colab which is a browser based programming environment so I can show you how this works. And another reason we don't need to use

1:18machine learning is that a rule-based solution is exact and requires no learning. So this is not learning this is just doing what we tell it to do and using machine learning would add complexity and reduce accuracy. So reducing complexity is key. We want to make sure that we don't do that and that's going to be a recurring theme in this course. So when would you use machine learning? So you would use it when the patterns must be inferred from data not when the exact answers can be computed directly as we see here. So we know that software 1.0 is deterministic. If X then Y and we can call this a rule-based system. But when you have a problem or let's say an AI problem and you're trying to solve it how do you choose? You start with the simplest tool that works.

2:10If a rule-based system like software 1.0 works then you ship 1.0. If you need prediction from structured data then you might consider ML but what you're doing there is you're adding complexity. If the signal lives in unstructured content then you might move to deep learning and that you add more complexity. Maybe you need fluent text or code or summaries then you might use Gen AI. And they're all useful but you would add more and more complexity only if you need it because software 1.0 didn't work and then here software 2.0 didn't work and then you would add software 3.0. So as you go higher in the software scale you're adding more complexity. So let's try out some examples here. Let me make this a little bit bigger and this is Google Colab. So you can just start to write code. You can say 1 plus 1 and then you can run that and immediately once it connects to the cloud it'll give you that answer. But we could also say something like text. There's an answer 2. 1 plus 1 is 2. That's a rule-based system. It's math but it's the same idea. But let's explore this. So let's say we have text equals and I'm going to use an input function and what this does is ask me a question. So I'm going to say what's your name and it's going to store

3:28my name here but I wanted to do something. So I wanted to return that name. So I'm going to say print name or text in this case but I want it to be uppercase. Alright so when I do this it's going to ask me my name. I'm going to do it in all lowercase and then enter and it's going to say Jonathan in uppercase. This is a series of rules. It's not learning anything. If I do it again and I say Jonathan it'll do it again. It doesn't matter what I do. If I do uppercase it's still going to give me uppercase. So this is rule-based. But how about the six-sided die? So I'm going to say total. So total outcomes is six. We have a six-sided fair die and the desired outcome. We want to know what the probability of rolling a four is and we're going to say there's one desired outcome out of six and you can calculate the probability by dividing the desired outcome by the total outcomes. Okay so now all I have to do is just print the probability. If I run this I'm going to get sixteen point seven percent and if I run it again I'm still going to get the same answer no matter how many times that I do this because it's a deterministic rule-based answer. And I'm going to share this notebook with you so that you can try it yourself and just become familiar that this is, let me go up here, what's cool about these notebooks

4:43you can add text. So I'm going to say I'm going to double click it I'm going to say software 1.0. Okay so this is example of software 1.0. In the next video we're going to explore software 2.0 but instead of programming it we're going to use these animals here. Well a hot dog and a hamburger and we're going to use a computer vision model. And normally when I teach courses about deep learning I would write out the code on how to do that. We're not going to do that. I'm going to show you a fun way that you can build a model yourself and so you can see the classification of each one of these objects yourself. We're going to do that with Teachable Machine. See you there.

0:00Welcome back. This is the fun part. We're going to use the hot dog and the hamburger to build our own computer vision model. So it's going to classify whether or not it's seeing, so we're teaching the computer how to see, if it's a hot dog or a hamburger. Normally, there would be a lot of code, and it'd be complicated. But what we're doing is really a no-code version of this, to give you the same output, and that's why it's fun.

0:30So you can understand, okay, I have a good idea of software 2.0. Let's check it out. All right, so here is Teachable Machines. Let's just click on Get Started, and we're going to use an image project. But what's interesting to know is that you could also use audio or a pose. So really, the image project and the pose project are very similar, because it's using the webcam for both of these. You could just use images as well, but we're going to use our webcam. So they're really similar. This one, the audio project, that's very different, but they're all the same 2.0 idea.

1:06And this is all deep learning. So let's click on Image Project. We're going to use the standard image model. And here, we're just going to name this Hot Dog and Hamburger. All right, so now we need to give it samples, and this is how it learns. This is the amazing part. So let me go ahead and click on the webcam. Okay, so here we are. I'm going to get out of the frame, and I'm going to make this a little bit smaller, something like that. So I'm going to crop the image a little bit, and I'm going to get out of the way. And the way we do this is we're going to hold the image up and move it around into different perspectives, all right, to give it training data.

1:41So it's learning to identify this by looking at the training data. So let's go ahead and hold record, and let's aim for about 100 samples, maybe 200. Let's do 200. It goes pretty quickly. Let's do 300. Okay, 310. That's good enough. All right, so let's turn this off and now go to the webcam here, and I'm going to crop it again over here. There we go. Done cropping, and now we're going to use the hamburger. So this one's a little more challenging. I have to move it in different angles. So 310.

2:12We want to have the same number of samples. That way the deep learning learns better. It learns faster. So I'm going to twist it around top, bottom. There's a magnet there. It's not part of a normal hamburger, but let's go ahead and do a little bit more. There, 309. That's close enough. So this is how the deep learning model learns. It's going to look at all of these images and understand what it's looking at. So let me click on this, and now it's preparing the training data. This is stuff that we would have to do programmatically.

2:43This is most of the code, and then it starts to train. Now you can see that it's going through different epochs. This is how many iterations. It does it over and over and over, and it passes all of that information through neurons in the artificial neural network. Okay, so now, moment of truth. I'm going to crop this as well and go over here. All right, so now I'm going to show it. That's, okay, now it's saying hamburger. And actually what I'm going to do here is I'm going to go back over here at a class, and I'm going to say nothing.

3:16So when we're looking at nothing, it knows that. So I'm going to say, I'm going to go to the webcam. I'm going to crop it just like we did before. Something like this, and I'm going to do 300 samples of nothing. And that's really important because otherwise it's predicting that it's looking at a hamburger, and it's not. So really, this is the nuts and bolts of machine learning and deep learning in this case, when you're talking about preparing the data. So now let's train it again. All right, so now it's saying nothing, 100%. You can see that down here. So it's working really well.

3:46Now I'm going to take the hot dog, and you can see that it says 100% hot dog. No matter which angle that I put it in, it knows that that's a hot dog. Let's try it all the way in the back. It even knows, look at that. It's performing really well. So now it's looking at nothing. Let's try hamburger. There we go, 100%, it's a hamburger. So our model is doing pretty good now that we added nothing. And this is a great example of our neural network classifier for computer vision. Now it's classifying hot dogs and hamburgers.

4:17But if I get in there, it thinks that I'm a mixture of all three. It thinks I'm mostly a hot dog, but I'm not. So that's something to consider. When you give it some other information, it starts to hallucinate. And by hallucination, the best example I can give you is when we get into software 3.0 in the next video. When you ask ChatGPT something that it doesn't know, it's going to guess, but it's going to do so with so much confidence that it's going to throw you off or anybody, really.

4:48It does it in a way that it's like, well, of course it must be true. It's so confident. But as you saw earlier, did you see all the different probabilities of me being a hot dog or maybe a hamburger or nothing? It was confused. That's when it would guess. So that's the danger that we need to look out for. And that's why there's so much guardrails that are built into these models and so much safety that goes into this when you're trying to apply this to a business situation. All right, I'll see you in the next video when we explore software 3.0 or generative AI.

5:23See you there.

0:00Welcome back. Here we're talking about generative AI, and it's a subset of deep learning. So here we have built on deep learning. So that's what it is. So again, AI is the umbrella term, and then you would have machine learning. And then the subset of that would be deep learning. And then the subset of that would be gen AI. We don't really talk about rule based AI, because that's the older version of AI. But you do know that it exists. But for the context of this course for executives and leaders, this is really all you need to know. And more importantly, software 1.0, 2.0, 3.0, because that's how you're going to choose how to solve different business problems. So what is generative AI? So let's say you have text, some kind of historical data, videos, or unlabeled data, then you would pass that into a model. And you see where it says pre training, we just did training ourselves. So we took, in our case, video feed, and then we trained our model. And then it was able to do we could say information extraction, we can call it computer vision, but that's what it did. But we didn't use plain language. So when we're talking about adapting, we're not going to be using lines of code like we did before, or even teachable

1:15machine. Normally, when you're doing software 2.0, you're writing code, but the model learns by itself, we're not telling it what to do, like you would in software 1.0. And the final part here is software 3.0 is that we're writing English, we're just telling it what we want. And that's the adapting part, right? Once you train that model, you can just talk to it in plain language. And it gives you the output much like we did with software 2.0. Let me prove it to you. It's really interesting. Okay, so let's take software 2.0, and software 3.0 and compare it. So first, I gave it some input data, which was a hamburger, or 300 images of hamburgers, and then 300 images of hot dogs, then we trained the model. And then it made a prediction, we didn't say here's the shape of a hamburger, the difference between a hamburger and a hot dog is that one is going to be longer like this and have a bun with, you know, ketchup, mustard and things like that. Hamburgers more like round shapes with a patty and then lettuce and stuff. We didn't say any of that. It learned it by itself during training, right? Software 1.0, you would have to explain all of that, but not software 2.0. And then when it made the prediction, it gave us a probability. In our case, when we

2:34showed it a picture of this hot dog, it said that it was a 100%. That's a hot dog. So very high degree of accuracy. With software 3.0, or Gen AI, let me put machine learning and deep learning over here. But really, we were doing deep learning, which is better for more complex data. So here's how it works. We give it input data. So I'm going to give it a picture of a hot dog. And then it's just going to give me an output, it's going to give us text, human, human like text, in fact. So the main difference is that we didn't give a bunch of these images to train, we don't have to train the model. But when we tested our teachable machine, we held something up, right? What is this hamburger 100%? Same thing happens with software 3.0. If I were to give it an image like this, and I would ask it, what is it, it would reply in human like text, say that's a hamburger, and it would describe the hamburger, it's more complex. And we're not training one specific thing.

3:35Software 3.0 has been trained on a gigantic amount of data, in this case, most of the internet, and it's able to generalize. So if you give it almost anything you can imagine that you have a picture of, it would do a pretty good job of telling you what it is, right? As long as it's been trained on that data. So let's take a look. So here we have chat GPT. And that just gave it text, I told it what I wanted. And it gave me this nebula painting of basketball player, right? And here I give it a really detailed prompt. And again, this has to do with prompt engineering, if you say give me a dinosaur image, it's not going to be great. But when you're really leveraging prompt engineering, you can get a great image like this. But let's start a new chat. Okay, so here we have an image of a hot dog. This is the exact hot dog that we used for software 2.0. But we're not using it to train, we just want to know what is this. So I've asked the question, what is the probability of the class of this image? Before we gave it the class, we said these are 300 pictures of a hot dog 300 pictures of nothing 300 pictures of a hamburger. Okay, so now look at this, it goes really into detail. So this is software 3.0, which is a subset of deep learning. So we're

4:54not training the models custom, this can just do almost anything that you ask of it. So now I'm going to ask for a percentage like we got it before. And here's an interesting let me make this much bigger. I don't have a classifier running yet. That's what we did in software 2.0. So I can't give you the actual probability values directly from a model, unless we set one up and run the image through it. That's what we did with TensorFlow inside of teachable machine. So saying use a pre trained image classification model, which is what we did, run the image through it, and then return the predicted top classes with their probability and percentage. Would you like me to run a plus hot dog image through one of these models and generate the probability percentages? Let's say yes. So instead of writing all the code, it'll just do it for us. But what we're doing is programming in plain language, because we're going to get the output. And that's where people don't really realize the power of software 3.0. They feel like they're talking to a human and the human is able to do stuff because they're a programmer and they can do all of this.

6:00Okay, but look what happened. Let's check it out. It looks like the classification pipeline failed because torch vision isn't fully supported in this environment. So it failed, it couldn't do it, it has limitations. So going back to this, we know that generative AI is built on top of deep learning, right? And it's trained on large scale data, meaning most of the internet, but you can adapt it to specific tasks using fine tuning or prompt engineering. So or prompting for short, so you can be very specific and leverage prompt engineering to adapt it to a specific task. And that's where you might not think that you have that power, because you feel like you're talking to a person. So when two humans are communicating, they understand the context, they understand guard rails, they understand all of these things. So you don't need to learn prompt engineering when speaking English. However, with software 3.0, prompt engineering is critical, more so when you're using it in a business. So when you're automating a task, for example, onboarding inside of your organization, you absolutely must consider prompt engineering. Furthermore,

7:16you might want to have a human in the loop. So they can look at the answer and be like, that is not appropriate. I'm so glad I caught this before we sent this to our new hire. Sometimes it can make mistakes that are catastrophic. And it's because of these reasons. So it's a general model. It's not a fine tune. So now that you know software 1.0, 2.0 and 3.0, let's review all of the different types of AI to make sure that we have this foundation down before we go on to the next skill. This is the takeaway. When we're thinking about artificial intelligence, this is the umbrella term for machine learning and its subset deep learning and its other subset generative AI. These two are software 2.0 and this is software 3.0 and software 1.0 is traditional rules. We tell it what to do. So in a way, software 1.0 is where we tell the computer what to do. Software 2.0, the computer learns on its own. After we build the model, we just give it the data and it learns from the data. Software 3.0, we're using plain language. Knowing each one of these types of software and each type of AI is crucial. There

8:32are other terms that you might hear that are not particularly important right now for our context, such as NLP or LLMs. NLP is natural language processing. That's how we teach computers to read and write. It's text-based. LLMs are large language models like ChatGPT, Gemini and Claude. All right, so that's it for this video and I'll see you in the challenge next. But just remember, there are three types of software and there are three subsets of AI and that's all we need before we get into the next skill. And the challenge is going to make sure that we have these concepts down before we move on to the next skill. All right, until next time, I hope this has been informative and I'd like to thank you for viewing.