1 00:00:00,080 --> 00:00:03,700 Everybody's talking about artificial intelligence these days, AI. 2 00:00:04,660 --> 00:00:07,019 Machine learning is another hot topic. 3 00:00:07,580 --> 00:00:09,779 Are they the same thing or are they different? 4 00:00:10,140 --> 00:00:12,119 And if so, what are those differences? 5 00:00:12,880 --> 00:00:15,500 And deep learning is another one that comes into play. 6 00:00:15,919 --> 00:00:21,100 I actually did a video on these three, artificial intelligence, machine learning, and deep learning, 7 00:00:21,219 --> 00:00:23,100 and talked about where they fit. 8 00:00:23,480 --> 00:00:26,239 And there were a lot of comments on that, and I read those comments, 9 00:00:26,239 --> 00:00:29,239 and I'd like to address some of the most frequently asked questions 10 00:00:29,239 --> 00:00:32,740 so that we can clear up some of the myths and misconceptions around this. 11 00:00:33,259 --> 00:00:37,020 In addition, something else has happened since that video was recorded, 12 00:00:37,159 --> 00:00:42,759 and that is the absolute explosion of this area of generative AI. 13 00:00:43,619 --> 00:00:49,340 Things like large language models and chatbots seem to be taking over the world. 14 00:00:49,640 --> 00:00:50,799 We see them everywhere. 15 00:00:51,359 --> 00:00:52,880 Really interesting technology. 16 00:00:53,299 --> 00:00:55,799 And then also things like deepfakes. 17 00:00:56,240 --> 00:01:01,460 These are all within the realm of AI, but how do they fit within each other? 18 00:01:01,799 --> 00:01:03,340 How are they related to each other? 19 00:01:03,640 --> 00:01:10,780 We're going to take a look at that in this video and try to explain how all these technologies relate and how we can use them. 20 00:01:11,120 --> 00:01:12,659 First off, a little bit of a disclaimer. 21 00:01:12,900 --> 00:01:19,000 I'm going to have to simplify some of these concepts in order to not make this video last for a week. 22 00:01:19,000 --> 00:01:23,519 So those of you that are really deep experts in the field, apologies in advance. 23 00:01:23,519 --> 00:01:27,719 But we're going to try to make this simple, and that will involve some generalizations. 24 00:01:27,980 --> 00:01:29,900 First of all, let's start with AI. 25 00:01:30,579 --> 00:01:40,579 Artificial intelligence is basically trying to simulate with a computer something that would match or exceed human intelligence. 26 00:01:41,000 --> 00:01:41,799 What is intelligence? 27 00:01:42,200 --> 00:01:49,900 Well, it could be a lot of different things, but generally we tend to think of it as the ability to learn, to infer, and to reason, things like that. 28 00:01:49,900 --> 00:01:55,680 So that's what we're trying to do in the broad field of AI, of artificial intelligence. 29 00:01:55,939 --> 00:02:00,980 And if we look at a timeline of AI, it really kind of started back around this time frame. 30 00:02:01,200 --> 00:02:04,140 And in those days, it was very premature. 31 00:02:04,659 --> 00:02:06,140 Most people had not even heard of it. 32 00:02:06,540 --> 00:02:09,379 And it basically was a research project. 33 00:02:09,759 --> 00:02:16,900 But I can tell you, as an undergrad, which for me was back during these times, we were doing AI work. 34 00:02:16,900 --> 00:02:22,319 In fact, we would use programming languages like Lisp or Prolog. 35 00:02:22,860 --> 00:02:29,460 And these kinds of things were kind of the predecessors to what became later expert systems. 36 00:02:30,120 --> 00:02:38,659 And this was a technology, again, some of these things existed previous, but that's when it really hit kind of a critical mass and became more popularized. 37 00:02:38,960 --> 00:02:42,539 So expert systems of the 1980s, maybe in the 90s. 38 00:02:42,840 --> 00:02:45,020 And again, we use technologies like this. 39 00:02:45,020 --> 00:02:52,120 all of this was something that we did before we ever touched in to the next topic I'm going to 40 00:02:52,120 --> 00:02:59,319 talk about. And that's the area of machine learning. Machine learning is, as its name implies, 41 00:02:59,639 --> 00:03:04,639 the machine is learning. I don't have to program it. I give it lots of information and it observes 42 00:03:04,639 --> 00:03:10,280 things. So for instance, if I start doing this, if I give you this and then ask you to predict 43 00:03:10,280 --> 00:03:14,180 what's the next thing that's going to be there, well, you might get it, you might not. You have 44 00:03:14,180 --> 00:03:19,639 very limited training data to base this on. But if I gave you one of those and then ask you what 45 00:03:19,639 --> 00:03:23,500 to predict would happen next, well, you're probably going to say this. And then you're 46 00:03:23,500 --> 00:03:27,300 going to say it's this. And then you think you got it all figured out. And then you see one of 47 00:03:27,300 --> 00:03:32,500 these. And then all of a sudden, I give you one of those and throw you a curveball. So this, 48 00:03:32,800 --> 00:03:38,400 in fact, and then maybe it goes on like this. So a machine learning algorithm is really good 49 00:03:38,400 --> 00:03:43,659 at looking at patterns and discovering patterns within data. The more training data you can give 50 00:03:43,659 --> 00:03:49,460 the more confident it can be in predicting. So predictions are one of the things that machine 51 00:03:49,460 --> 00:03:55,639 learning is particularly good at. Another thing is spotting outliers like this and saying, 52 00:03:55,840 --> 00:04:00,319 oh, that doesn't belong in, it looks different than all the other stuff because the sequence 53 00:04:00,319 --> 00:04:05,860 was broken. So that's particularly useful in cybersecurity, the area that I work in, 54 00:04:05,960 --> 00:04:10,379 because we're looking for outliers. We're looking for users who are using the system in ways that 55 00:04:10,379 --> 00:04:15,719 they shouldn't be or ways that they don't typically do. So this technology, machine learning, 56 00:04:15,759 --> 00:04:21,000 is particularly useful for us. And machine learning really came along and became more 57 00:04:21,000 --> 00:04:29,720 popularized in this time frame, in the 2010s. And again, back when I was an undergrad writing my 58 00:04:29,720 --> 00:04:36,500 class, we were doing this kind of stuff. We never once talked about machine learning. It might have 59 00:04:36,500 --> 00:04:43,019 existed, but it really hadn't hit the popular mindset yet. But this technology has matured 60 00:04:43,019 --> 00:04:48,339 greatly over the last few decades, and now it becomes the basis of a lot we do going forward. 61 00:04:48,779 --> 00:04:55,420 The next layer of our Venn diagram involves deep learning. Well, it's deep learning in the sense 62 00:04:55,420 --> 00:05:01,939 that with deep learning, we use these things called neural networks. Neural networks are ways 63 00:05:01,939 --> 00:05:07,220 that in a computer we simulate and mimic the way the human brain works, at least to the extent that 64 00:05:07,220 --> 00:05:11,879 we understand how the brain works. And it's called deep because we have multiple layers of those 65 00:05:11,879 --> 00:05:18,240 neural networks. And the interesting thing about these is they will simulate the way a brain 66 00:05:18,240 --> 00:05:23,000 operates. But I don't know if you've noticed, but human brains can be a little bit unpredictable. 67 00:05:23,379 --> 00:05:28,379 You put certain things in, you don't always get the very same thing out. And deep learning is the 68 00:05:28,379 --> 00:05:33,959 same way. In some cases, we're not actually able to fully understand why we get the results we do 69 00:05:33,959 --> 00:05:39,379 because there are so many layers to the neural network, it's a little bit hard to decompose 70 00:05:39,379 --> 00:05:44,480 and figure out exactly what's in there. But this has become a very important part and a very 71 00:05:44,480 --> 00:05:51,560 important advancement that also reached some popularity during the 2010s and as something 72 00:05:51,560 --> 00:05:58,079 that we use still today as the basis for our next area of AI. The most recent advancements 73 00:05:58,079 --> 00:06:03,920 in the field of artificial intelligence, all really are in this space, the area of generative 74 00:06:03,920 --> 00:06:08,480 AI. Now, I'm going to introduce a term that you may not be familiar with. It's the idea of 75 00:06:08,480 --> 00:06:14,680 foundation models. Foundation models is where we get some of these kinds of things. For instance, 76 00:06:14,819 --> 00:06:21,180 an example of a foundation model would be a large language model, which is where we take language 77 00:06:21,180 --> 00:06:26,699 language and we model it and we make predictions in this technology where if I see certain 78 00:06:26,699 --> 00:06:31,699 types of words, then I can sort of predict what the next set of words will be. 79 00:06:31,699 --> 00:06:35,959 I'm going to oversimplify here for the sake of simplicity, but think about this as a little 80 00:06:35,959 --> 00:06:38,360 bit like the autocomplete. 81 00:06:38,360 --> 00:06:43,279 When you start typing something in and then it predicts what your next word will be, except 82 00:06:43,279 --> 00:06:47,759 in this case with large language models, they're not predicting the next word, they're predicting 83 00:06:47,759 --> 00:06:51,579 the next sentence, the next paragraph, the next entire document. 84 00:06:51,920 --> 00:06:56,560 So there's really an amazing exponential leap in what these things are able to do. 85 00:06:56,879 --> 00:07:04,240 And we call all of these technologies generative because they are generating new content. 86 00:07:05,079 --> 00:07:09,800 Some people have actually made the argument that the generative AI isn't really generative, 87 00:07:09,800 --> 00:07:15,360 that these technologies are really just regurgitating existing information and putting it in different format. 88 00:07:15,740 --> 00:07:17,259 Well, let me give you an analogy. 89 00:07:17,759 --> 00:07:23,560 If you take music, for instance, then every note has already been invented. 90 00:07:24,240 --> 00:07:33,579 So, in a sense, every song is just a recombination, some other permutation of all the notes that already exist already and just putting them in a different order. 91 00:07:34,079 --> 00:07:37,139 Well, we don't say new music doesn't exist. 92 00:07:37,620 --> 00:07:42,480 People are still composing and creating new songs from the existing information. 93 00:07:42,939 --> 00:07:44,779 I'm going to say Gen AI is similar. 94 00:07:45,180 --> 00:07:49,160 It's an analogy, so there'll be some imperfections in it, but you get the general idea. 95 00:07:49,740 --> 00:07:55,220 Actually, new content can be generated out of these, and there are a lot of different forms that this can take. 96 00:07:55,220 --> 00:08:02,759 Other types of models are audio models, video models, and things like that. 97 00:08:03,060 --> 00:08:07,240 Well, in fact, these we can use to create deepfakes. 98 00:08:08,019 --> 00:08:15,079 And deep fakes are examples where we're able to take, for instance, a person's voice and recreate that 99 00:08:15,079 --> 00:08:18,600 and then have it seem like the person said things they never said. 100 00:08:18,879 --> 00:08:23,959 Well, it's really useful in entertainment situations, in parodies and things like that, 101 00:08:24,180 --> 00:08:27,220 or if someone's losing their voice, then you could capture their voice, 102 00:08:27,339 --> 00:08:30,420 and then they'd be able to type and you'd be able to hear it in their voice. 103 00:08:30,600 --> 00:08:33,720 But there's also a lot of cases where this stuff could be abused. 104 00:08:34,299 --> 00:08:37,779 The chatbots, again, come from this space. 105 00:08:38,259 --> 00:08:41,039 The deepfakes come from this space. 106 00:08:41,120 --> 00:08:45,779 But they're all part of generative AI and all part of these foundation models. 107 00:08:46,279 --> 00:08:52,200 And this, again, is the area that has really caused all of us to really pay attention to AI. 108 00:08:52,779 --> 00:08:58,159 The possibilities of generating new content or, in some cases, summarizing existing content 109 00:08:58,159 --> 00:09:01,980 and giving us something that is bite-sized and manageable. 110 00:09:02,580 --> 00:09:04,700 This is what has gotten all of the attention. 111 00:09:05,240 --> 00:09:07,700 This is where the chatbots and all of these things come in. 112 00:09:08,360 --> 00:09:12,080 In the early days, AI's adoption started off pretty slowly. 113 00:09:12,820 --> 00:09:17,919 Most people didn't even know it existed, and if they did, it was something that always seemed like it was about 5 to 10 years away. 114 00:09:18,399 --> 00:09:23,659 But then machine learning, deep learning, and things like that came along, and we started seeing some uptick. 115 00:09:24,259 --> 00:09:29,440 Then foundation models, Gen AI, and the light came along, and this stuff went straight to the moon. 116 00:09:29,440 --> 00:09:37,440 These foundation models are what have changed the adoption curve and now you see AI being adopted everywhere. 117 00:09:37,440 --> 00:09:45,440 And the thing for us to understand is where this is, where it fits in, and make sure that we can reap the benefits from all of this technology. 118 00:09:45,440 --> 00:09:50,440 If you like this video and want to see more like it, please like and subscribe. 119 00:09:50,440 --> 00:09:55,440 If you have any questions or want to share your thoughts about this topic, please leave a comment below.