#73 – Andrew Ng: Deep Learning, Education, and Real-World AI

Growing up in Hong Kong, Singapore, I started learning to code when I was five or six years old. At that time, I was learning the BASIC programming language, and they would take these books and you know they'll tell you type this program into your computer. I'd type that program on my computer, and as a result of all that typing, I would get to play these very simple shoot-em-up games that you know I had implemented on my little computer. So I thought it was fascinating as a young kid that I could write this code that was really just copying code from a book into my computer to then play these cool little video games. Another moment for me was when I was a teenager, and my father, who's a doctor, was reading about expert systems and about neural networks. So he got me to read some of these books, and I thought it was really cool you could write a computer that started to exhibit intelligence. Then I remember doing an internship while I was in high school. This is in Singapore, where I remember doing a lot of photocopying and as office assistant. And the highlight of my job was when I got to use the shredder. So the teenager of me remember thinking, boy, this is a lot of photocopying. If only we could write software, build a robot, something to automate this, maybe I could do something else. So I think a lot of my work since then has centered on the theme of automation. Even the way I think about machine learning today, we're very good at writing learning algorithms that can automate things that people can do. Or even launching the first MOOCs, massive open online courses, that later led to Coursera. I was trying to automate what could be automatable in how I was teaching on campus, the process of education, tried to automate parts of that to make it more sort of to have more impact from a single teacher, single educator.

You know, teaching online, what not many people know was that a lot of those videos were shot between the hours of 10:00 p.m. and 3:00 a.m. A lot of times, we were launching the first Stanford course. We already announced the course, about 100,000 people signed up. We just started to write the code, and we had not yet actually filmed the video. So, we had a lot of pressure, 100,000 people waiting for us to produce the content. So many Fridays, Saturdays, I would go out, have dinner with my friends, and then I would think, okay, do you want to go home now or do you want to go to the office to film videos? And the thought of being able to help 100,000 people potentially learn machine learning, fortunately that made me think, okay, I want to go to my office, go to my tiny little recording studio. I would adjust my Logitech webcam, adjust my Wacom tablet, make sure my lapel mic was on, and then I would start recording often until 2:00 a.m. or 3:00 a.m. I think it's unfortunate that it doesn't show that it was recorded that late at night, but it was really inspiring the thought that we could create content to help so many people learn about machine learning.

I think it's humbling, and I wasn't thinking about what I was feeling. I think one thing that I'm proud to say we got right from the early days was I told my whole team back then that the number one priority is to do what's best for learners, do what's best for students. And so when I went to the recording studio, the only thing on my mind was what can I say? How can I design my slides? What do I need to draw right to make these concepts as clear as possible for learners. I think you know I've seen sometimes instructors, it's tempting to hey let's talk about my work, maybe if I teach you about my research someone will cite my papers a couple more times. And I think one of the things we got right launching the first few MOOCs and later building Coursera was putting in place that bedrock principle of let's just do what's best for learners and forget about everything else. And I think that as a guiding principle turned out to be really important to the rise of the MOOC movement.

I really want to help anyone that had an interest in machine learning to break into the field. And I think sometimes I've actually had people ask me, hey, why are you spending so much time explaining gradient descent? And my answer was if I look at what I think the learner needs and would benefit from, I felt that having a good understanding of the foundations, kind of back to the basics, would put them in a better state to then build on a long-term career. So tried to consistently make decisions on that principle.

I think the number has grown over time. I think it's one of those things that maybe it feels like it came out of nowhere, but as an insider building it, it took years. It's one of those overnight successes that took years to get there.

My first foray into this type of online education was when we were filming my Stanford class and sticking the videos on YouTube and some other things. We had uploaded a whole and so on. But you know, basically the one hour fifteen minute video that we put on YouTube, and then we had four or five other versions of websites that I had built, most of which you would never have heard of because they reached small audiences, but that allowed me to iterate, allow my team and me to iterate to learn what are the ideas that work and what doesn't. For example, one of the features I was really excited about and really proud of was building this website where multiple people could be logged into the website at the same time. So today if you go to a website, you know, if you're logged in and then I want to log in, you need to log out if it's the same browser, same computer. But I thought, well, what if two people, say you and me, we're watching a video together in front of a computer. What if a website could have you type your name and password, have me type my name and password, and then now the computer knows both of us are watching together and it gives both of us credit for anything we do as a group. I even featured this at a high school in San Francisco. We had about twenty-something users, where's the teacher there, Sacred Heart Cathedral Prep, the teacher is great. And guess what? Zero people used this feature. It turns out people studying online, they want to watch the videos by themselves. You can play back, pause at your own speed rather than in groups. So that was one example of a tiny lesson learned out of many that allowed us to hone in to the set of features.

And I think with the whole machine learning movement as well, I think it didn't come out of nowhere. Instead what happened was as more people learn about machine learning, they will tell their friends and their friends will see how it's applicable to their work and then the community kept on growing. And I think we're still growing. You know, I don't know in the future what percentage of all developers will be AI developers. I could easily see it being north of 50%. Right? Because so many AI developers broadly construed, not just people doing the machine learning modeling but the people building infrastructure, data pipelines, you know, all the software surrounding the core machine learning model, maybe is even bigger. I feel like today almost every software engineer has some understanding of the cloud. Not all, you know, but maybe this microcontroller developer doesn't need to deal with the cloud. But I feel like the vast majority of software engineers today sort of have an appreciation of cloud. I think in the future maybe we'll approach nearly 100% of all developers being you know in some way an AI developer or at least having an appreciation of machine learning. And my hope is that there's this kind of effect that there's people who are not really interested in being a programmer or being into software engineering, like biologists, chemists, and physicists, even mechanical engineers, all these disciplines that are now more and more sitting on large data sets. And here they didn't think they're interested in programming until they have this data set and they realize there's these set of machine learning tools that allow you to use the data set. So they actually become, they learn to program and they become new programmers. So not just because you've mentioned a larger percentage of developers become machine learning people, it seems like more and more the kinds of people who are becoming developers is also growing significantly.

Yeah. I think once upon a time only a small part of humanity was literate, you could read and write, and maybe you thought maybe not everyone needs to learn to read and write, you just go listen to a few monks, right? Read to you. And maybe that was enough. Or maybe just need a few handful of authors to write the bestsellers and then no one else needs to write. But what we found was that by giving as many people, you know, in some countries almost everyone basic literacy, it dramatically enhanced human-to-human communications and we can now write for an audience of one such as if I send you an email or you send me an email. I think in computing we're still in that phase where so few people know how to code that the coders mostly have to code for relatively large audiences. But if everyone or most people became developers at some level, similar to how most people in developed economies are somewhat literate, I would love to see the owners of a mom and pop store be able to write a little bit of code to customize the TV display for their special this week. Yeah, I think it'll enhance human-to-computer communications which is becoming more and more important today as well.

I could see society getting there. There's one other interesting thing, you know, if I go talk to the mom and pop store, if I talk to a lot of people in their daily professions, I previously didn't have a good story for why they should learn to code, you know, we could give them some reasons but what I found with the rise of machine learning and data science is that I think the number of people with a concrete use for data science in their daily lives, in their jobs, may be even larger than the number of people with a concrete use for software engineering. For example, if you run a small mom and pop store, I think if you can analyze the data about your sales, your customers, I think there's actually real value there, maybe even more than traditional software engineering. So I find that for a lot of my friends in various professions, be it recruiters or accountants or you know people that work in factories which I deal with more and more these days, I feel if they were data scientists at some level they could immediately use that in their work. So I think that data science and machine learning may be an even easier entry into the developer world for a lot of people than software engineering.

I think it depends on the concepts you want to explain. For mathematical concepts it's nice you can build up the equation one piece at a time. And the whiteboard marker or the pen and stylus is a very easy way to build up an equation, build up a complex concept one piece at a time while you're talking about it, and sometimes that enhances understandability. The downside of writing is that it's slow and so if you want a long sentence it's very hard to write that so I think there are pros and cons and sometimes I use slides and sometimes I use a whiteboard or a stylus. The slowness of a whiteboard is also its upside because it forces you to reduce everything to the basics.

Yeah. I was really fortunate to have had Pieter Abbeel as my first PhD student. And I think even my long-term professional success builds on early foundations or early work that Pieter was so critical to. So, it was really grateful to him for working with me. You know, what not a lot of people know is just how hard research was and still is. Pieter's PhD thesis was using reinforcement learning to fly helicopters. And so, you know, actually even today the website heli.stanford.edu is still up. You can watch videos of us using reinforcement learning to make a helicopter fly upside down, fly loops, roses. It's cool.

Oh, yeah. Thanks.

Something like that. It's like over 10 years old.

Yes. What not many people see is how hard it was. So Pieter and Adam Coates and Morgan Quigley and I were working on various versions of the helicopter and a lot of things did not work. For example, turns out one of the hardest problems we had was when helicopters flying around upside down doing stunts, how do you figure out the position? How do you localize a helicopter? So we wanted to try all sorts of things. Having one GPS unit doesn't work because you're flying upside down. GPS unit is facing down so you can't see the satellites. So, we experimented trying to have two GPS units, one facing up, one facing down. So, if you flip over, that didn't work because the downward facing one couldn't synchronize if you're flipping quickly. Morgan Quigley was exploring this crazy complicated configuration of specialized hardware to interpret GPS signals, looking into FPGA, completely insane. Spent about a year working on that. Didn't work. So I remember Pieter, great guy, him and me, you know, sitting down in my office looking at some of the latest things we had tried that didn't work and saying, you know, darn it, like what now? Because we tried so many things and it just didn't work. In the end, what we did, and Adam Coates was crucial to this, was put cameras on the ground and use cameras on the ground to localize a helicopter, and that solved the localization problem, so that we could then focus on the reinforcement learning and inverse reinforcement learning techniques to then actually make the helicopter fly. And you know I'm reminded when I was doing this work at Stanford around that time there was a lot of reinforcement learning theoretical papers but not a lot of practical applications. So the autonomous helicopter work for flying helicopters was one of the few practical applications of reinforcement learning at the time which caused it to become pretty well-known. I feel like we might have almost come full circle where today there's so much buzz, so much hype, so much excitement about reinforcement learning, but again we're hunting for more applications of all of these great ideas that the community's come up with.

I like stuff that works. In the physical world. So like it's back to the shredder. You know, I like theory, but when I work on theory myself, and this is personal taste. I'm not saying anyone else should do what I do, but when I work on theory, I personally enjoy it more if I feel that the work I do will influence people, have positive impact or help someone. I remember when many years ago, I was speaking with a mathematics professor and I just said, hey, why do you do what you do? And he actually, you know, he had stars in his eyes when he answered, and this mathematician, not from Stanford, different university, he said I do what I do because it helps me to discover truth and beauty in the universe.

And I thought that's great. I don't want to do that. I think it's great that someone does that, fully support the people that do it, a lot of respect for people for that. But I am more motivated when I can see a line to how the work that my teams and I are doing helps people. The world needs all sorts of people. I'm just one type. I don't think everyone should do things the same way as I do. But when I delve into either theory or practice, if I personally have conviction, you know, that here's a pathway to help people, I find that more satisfying to have that conviction.

Yeah I can tell you the thing we got wrong and the thing we got right. The thing we really got wrong was the importance of, the early importance of unsupervised learning. So early days of Google Brain we put a lot of effort into unsupervised learning rather than supervised learning. And there was this argument actually I think it was around 2005 after, you know, NeurIPS at that time called NIPS but now NeurIPS had ended and Jeff Hinton and I were sitting in the cafeteria outside, you know, the conference we had lunch, we were just chatting, and Jeff pulled up this napkin, he started sketching this argument on a napkin. It was very compelling. I'll repeat it. Human brain has about 100 trillion, so that's 10 to the 14, synaptic connections. You will live for about 10 to the 9 seconds, that's 30 years. You actually live for two by 10 to the 9, maybe three by 10 to the 9 seconds. So just let's say 10 to the 9. So if each synaptic connection, each weight in your brain, your network, has just a one bit parameter, that's 10 to the 14 bits you need to learn in up to 10 to the 9 seconds of your life. So via this simple argument, which like a lot of problems is very simplified, that's 10 to the 5 bits per second you need to learn in your life. And I have a one-year-old daughter. I am not pointing out 10 to the 5 bits per second of labels to her. So, and I think I'm a very loving parent, but I'm just not going to do that, right? So from this, you know, very crude, definitely problematic argument, there's just no way that most of what we know is through supervised learning. But where you could get so many bits of information is from sucking in images, audio, just experiences in the world. And so that argument, and there are a lot of nuances to the argument, you know, we should go into, really convinced me that there's a lot of power to unsupervised learning. So that was the part that we actually maybe got wrong. I still think unsupervised learning is really important, but in the early days, you know, 10, 15 years ago, a lot of us thought that was the path forward.

For the time, that was the part we got wrong. The part we got right was the importance of scale. So Adam Coates, another wonderful person fortunate to have worked with him. He was in my group at Stanford at the time and Adam had run these experiments at Stanford showing that the bigger we trained a learning algorithm, the better his performance. And it was based on that, there was a graph that Adam generated, you know, where the x-axis, y-axis, lines going up and to the right, bigger make this thing the better performance, accuracy is a vertical axis. So it's really based on that chart that Adam generated that he gave me the conviction that we could scale these models way bigger than what we could on the CPUs, which is what we had at Stanford, that we could get even better results. And it was really based on that one figure that Adam generated that gave me the conviction to go with Sebastian Thrun to pitch, you know, starting a project at Google, which became the Google Brain project.

It was controversial at the time. Some of my well-meaning friends, you know, senior people in the machine learning community, I won't name, but people some of whom we know. My well-meaning friends came and were trying to give me friendly warnings like, hey Andrew why are you doing this? This is crazy. Look at these architectures are building. You just want to go for scale? Like this is a bad career move. So, so my well-meaning friends, you know, were trying, some of them were trying to talk me out of it. But I find that if you want to make a breakthrough, you sometimes have to have conviction and do something before it's popular, since that lets you have a bigger impact.

I think both are important and it's also problem dependent. So for a few data sets we may be approaching, you know, Bayes error rate, or approaching or surpassing human level performance, and then there's that theoretical ceiling that we will never surpass a Bayes error rate. But then I think there are plenty of problems where we're still quite far from either human level performance or from Bayes error rate, and bigger data sets with neural networks, without further algorithm innovation, will be sufficient to take us further. But on the flip side if we look at the recent breakthroughs using, you know, transformer networks or language models, it was a combination of novel architecture, but also scale had a lot to do with it. If we look at what happened with GPT-2 and BERT, I think scale was a large part of the story.

Right. I find that today we have maturing processes to managing code, things like Git, right, version control. It took us a long time to evolve the good processes. I remember when my friends and I were emailing each other C++ files in email, you know, but then we had, was it CVS, version control, get maybe something else in the future. We're very immature in terms of tools for managing data and think of how to clean data and how to solve very messy data problems. I think there's a lot of innovation there to be had still.

I'll give you one example. When we work with manufacturing companies, it's not at all uncommon for there to be multiple labelers that disagree with each other, right? And so doing the work in visual inspection, we would, you know, take say a plastic part and show it to one inspector, and the inspector, sometimes very opinionated, they go, clearly that's a defect, this scratch, unacceptable, got to reject this part. Take the same part to different inspector, different, very opinionated. Clearly the scratch is small, it's fine, don't throw it away, you're going to make us lose money. And then sometimes you take the same plastic part, show it to the same inspector in the afternoon as opposed to the morning, and very confidently, in the morning they say clearly it's okay, in the afternoon equally confident, clearly this is a defect. And so what is an AI team supposed to do if sometimes even one person doesn't agree with himself or herself in the span of a day? So I think these are the types of very practical, very messy data problems that my teams wrestle with. In the case of large consumer internet companies where you have a billion users, you have a lot of data, you don't worry about it. Just take the average, it kind of works. But in a case of other industry settings, we don't have big data. If just small data, very small data sets, maybe on a hundred defective parts or a hundred examples of a defect. If you have only a hundred examples, these little labeling errors, you know, if ten of your hundred labels are wrong, that actually is ten percent of your data set, has a big impact. So, how do you clean this up? What are you supposed to do? This is an example of the types of things that my teams, this is a Landing AI example, are wrestling with to deal with small data, which comes up all the time once you're outside consumer internet.

Yeah. I find it actually when I'm teaching at Stanford, I increasingly encourage students at Stanford to try to find their own project for the end of term project, rather than just downloading someone else's nicely clean data set. It's actually much harder if you need to go find your own problem and find your own data set rather than you go to one of the several good websites, very good websites with clean scoped data sets that you could just work on.

No. Yeah. Thank you. You know, I was once reading a news article, I think it was Tech Review, and I'm going to mess up the statistic, but I remember reading an article that said something like one-third of all programmers are self-taught. I may have the number one-third around, maybe it was two-thirds, but when I read that article, I thought this doesn't make sense. Everyone is self-taught. Because you teach yourself, I don't teach people. I just don't.

So the Deep Learning Specialization offered by DeepLearning.AI is, I think, one, it was Coursera's top specialization. It might still be. So it's a very popular way for people to take that specialization to learn about everything from neural networks to how to tune your network to what is a CNN, to what is an RNN or sequence model, or what is an attention model. And so the Deep Learning Specialization steps everyone through those algorithms so you deeply understand it and can implement it and use it for whatever application, from the very beginning.

Yeah, need to understand basic programming since there are programming exercises in Python. And the math prereq is quite basic. So no calculus is needed. If you know calculus, it's great, you get better intuitions. But I deliberately tried to teach that specialization without requiring calculus. So I think high school math would be sufficient. If you know how to multiply two matrices, I think that's great.

Basic linear algebra, even very, very basic linear algebra and some programming. I think that people that have done the machine learning course will find the Deep Learning Specialization a bit easier, but it's also possible to jump into the Deep Learning Specialization directly, but it'll be a little bit harder since we tend to, you know, go over faster concepts like how does gradient descent work and what is an objective function, which is covered more slowly in the machine learning course.

So if you take the Deep Learning Specialization, you learn the foundations of what is a neural network, how do you build up a neural network from a, you know, single logistic unit to a stack of layers to different activation functions, you learn how to train the neural networks. One thing I'm very proud of in that specialization is we go through a lot of practical know-how of how to actually make these things work. So what are the differences between different optimization algorithms? What do you do if the algorithm overfits or how do you tell if the algorithm is overfitting? When do you collect more data? When should you not bother to collect more data? I find that even today unfortunately there are engineers that will spend six months trying to pursue a particular direction such as collect more data because we heard more data is valuable. But sometimes you could run some tests and could have figured out six months earlier that for this particular problem collecting more data isn't going to cut it. So just don't spend six months collecting more data, spend your time modifying the architecture or trying something else. So it goes through a lot of the practical know-how so that when you take the Deep Learning Specialization you have those skills to be very efficient in how you build these networks, to dive right in to play with the network, to train it, to do the inference on a particular data set, to build the intuition about it without building it up too big to where you spend like you said six months learning, building up your big project without building any intuition of a small aspect of the data that could already tell you everything you need to know about that data.

And also the systematic frameworks of thinking for how to go about building practical machine learning. Maybe to make an analogy, when we learn to code, we have to learn the syntax of some programming language, right? Be it Python or C++ or Octave or whatever. But the equally important or maybe even more important part of coding is to understand how to string together these lines of code into coherent things. So you know when should you put something in a function call and when should you not? How do you think about abstraction? So those frameworks are what makes a programmer efficient even more than understanding the syntax. I remember when I was in undergrad at Carnegie Mellon, one of my friends would debug their code by first trying to compile it, and then it was C++ code, and then every line that is a syntax error, they want to get rid of syntax errors as quickly as possible. So how do you do that? Well, they would delete every single line of code with a syntax error. So really efficient for getting rid of syntax errors but horrible debugging services. So we learned how to debug. And I think in machine learning the way you debug a machine learning program is very different than the way you do binary search or whatever, use a debugger, trace through the code in traditional software engineering. So it is an evolving discipline but I find that the people that are really good at debugging machine learning algorithms are easily 10x, maybe 100x faster at getting something to work.

Yeah, often the question is why doesn't it work yet, or can I expect it to eventually work, and what are the things I could try: change the architecture, more data, more regularization, different optimization algorithm, you know, different types of data. So to answer those questions systematically so that you don't spend six months heading down a blind alley before someone comes and says why did you spend six months doing this.

Similar to learning mathematics, I think one of the challenges of deep learning is that there are a lot of concepts that build on top of each other. If you ask me what's hard about mathematics, I have a hard time pinpointing one thing. Is it addition, subtraction, is it a carry? Is it multiplication? There's just a lot of stuff. I think one of the challenges of learning math and of learning certain technical fields is that there are a lot of concepts, and if you miss a concept then you're kind of missing the prerequisite for something that comes later. So in the Deep Learning Specialization, we try to break down the concepts to maximize the odds of each component being understandable. So when you move on to the more advanced thing, we learn convolution, you have confidence, hopefully you have enough intuitions from the earlier sections to then understand why we structure convolution in a certain way and then eventually why we build, you know, RNNs and LSTMs or attention model in a certain way, building on top of the earlier concepts.

Well, I don't think anyone's ever turned the interview on me. I'm glad you go first.

I still teach reinforcement learning in one of my Stanford classes. And my PhD thesis was on reinforcement learning. So, I clearly love the field. I find that if I'm trying to teach students the most useful techniques for them to use today, I end up shrinking the amount of time I talk about reinforcement learning. It's not what's working today. Now, our work changes so fast. Maybe this will be totally different in a couple years. But I think we need a couple more things for reinforcement learning to get there.

One of my teams is looking at reinforcement learning for some robotic control tasks. So I see the applications but if you look at it as a percentage of all of the impact of the types of things we do, is at least today outside of playing video games, right, and a few other games, the scope actually at NeurIPS a bunch of us were standing around saying hey what's your best example of an actual deployed reinforcement learning application, and you know among senior machine learning researchers, right, and again there are some emerging ones but there are not that many great examples.

People have with reinforcement learning has been great, but not in the applied impact on the real world setting. So, it's a trade-off. How much impact you want to have versus how much fun you want to have.

Yeah, that's really cool. And I feel like, you know, the world actually needs all sorts. Even within machine learning, I feel like deep learning is so exciting, but AI team shouldn't just use deep learning. I find that my teams use a portfolio of tools. And maybe that's not the exciting thing to say, but some days we use a neural net. Some days we use PCA. Actually the other day I was sitting down with my team looking at PCA residuals trying to figure out what's going on with PCA applied to a manufacturing problem. And some days we use a probabilistic graphical model. Some days we use a knowledge graph, which is one of the things that has tremendous industry impact but the amount of chatter about knowledge graphs in academia is really thin compared to the actual raw impact. So, I think reinforcement learning should be in that portfolio and then it's about balancing how much we teach all of these things and the world should have diverse skills. It would be sad if, you know, everyone just learned one narrow thing.

I think that if my only job was being an academic researcher, if an unlimited budget and, you know, didn't have to worry about short-term impact and only focused on long-term impact, I probably spent all my time doing research on unsupervised learning. I still think unsupervised learning is a beautiful idea. At both this past NeurIPS and ICML, I was attending workshops or listening to various talks about self-supervised learning, which is one vertical segment maybe of sort of unsupervised learning that I'm excited about. Maybe just to summarize the idea, I guess, you know, the idea I'll describe briefly.

So here's an example of self-supervised learning. Let's say we grab a lot of unlabeled images off the internet. So we have infinite amounts of this type of data. I'm going to take each image and rotate it by a random multiple of 90 degrees. And then I'm going to train a supervised neural network to predict what was the original orientation. So has this been rotated 90, 180, 270, or 0 degrees. So you can generate an infinite amount of label data because you rotated the image so you know what's the ground truth label. And so various researchers have found that by taking unlabeled data and making up label data sets and training a large neural network on these tasks, you can then take the hidden layer representation and transfer it to a different task very powerfully. Learning word embeddings, where we take a sentence, delete a word, predict the missing word, which is how we learn, you know, one of the ways we learn word embeddings, is another example. And I think there's now this portfolio of techniques for generating these made-up tasks. Another one called jigsaw would be if you take an image, cut it up into a, you know, 3x3 grid. So like a nine 3x3 puzzle piece, jump out the nine pieces and have a neural network predict which of the nine factorial possible permutations it came from. So many groups, including, you know, OpenAI, FAIR has been doing some work on this too. Facebook, Google Brain, I think DeepMind, oh actually Aaron van den Oord has great work on the CPC objective. So many teams are doing exciting work and I think this is a way to generate infinite label data, and I find this a very exciting piece of unsupervised learning.

I don't think there's a whole enchilada. I think it's just a piece of it, and I think this one piece, these self-supervised learning, is starting to get traction. We're very close to it being useful. Well, word embeddings are really useful. I think we're getting closer and closer to this having a significant real world impact, maybe in computer vision and video. But I think this concept and I think there'll be other concepts around it. You know, other unsupervised learning things that I worked on, I've been excited about. I was really excited about sparse coding and slow feature analysis. I think all of these are ideas that various of us were working on about a decade ago before we all got distracted by how well supervised learning was working.

I think there's a lot more work that one could explore around this theme of ideas and other ideas to come up with better algorithms. So if we could return to, maybe talk quickly about the specifics of deep learning.ai, the deep learning specialization perhaps. How long does it take to complete the course, would you say?

The official length of the deep learning specialization is, I think, 16 weeks, so about 4 months, but it's go at your own pace. So if you subscribe to the deep learning specialization, there are people that finished it in less than a month by working more intensely and studying more intensely. It really depends on the individual. Yeah. When we created the deep learning specialization, we wanted to make it very accessible and very affordable. And with, you know, Coursera and deeplearning.ai education mission, one of the things that's really important to me is that if there's someone for whom paying anything is a financial hardship, then just apply for financial aid and get it for free.