Andrew Ng on winning the AI race, with DJ Patil

0:00If you've spent any time building AI products or leading technical teams, you know this. Transformation doesn't fail because of ideas. It fails because teams can't move together. Enter Atlassian's Teamwork Collection. It has planning in JIRA, documentation in Confluence, video updates in Loom, and now AI agents in Rovo, which connects the dots across your work so nothing gets lost. It's one AI-powered teamwork platform designed for how modern teams actually build. Learn more at Atlassian.com slash TeamChanger. That's A-T-L-A-S-S-I-A-N dot com slash teamchanger Starting a business comes with its share of ups and downs, which is why staying true to your vision is essential.

0:53A non-negotiable for Romeo and Milka Bregali, Capital One business customers and co-owners of Ra's plant-based restaurant in New York. Romeo and Milka took a leap of faith when starting their own restaurant, gutting an empty space and building it from the ground up. every pipe, every wall, every detail. But building from scratch came with a heavy financial burden, which is when they turned to their Capital One business card. With the flexibility of the card's no-pre-set spending limit, they were able to spend more and earn more rewards while bringing their vision to life. Today, Raz's success is proof that with passion and the right support, it's possible to make your dreams a reality.

1:30Learn more at CapitalOne.com slash business cards.

1:38Hi there, listeners. Today, we are sharing another conversation from this year's Masters of Scale Summit. Andrew Ng is a true AI pioneer. He's co-founder of Coursera and Deep Learning AI and managing partner at AI Fund, a studio that incubates new AI companies. He was joined on stage by DJ Patel, former chief data scientist under the Obama administration and general partner at Great Point Ventures. It's a dynamic discussion about the current state of AI, how future generations should approach the technology, and how America can stay competitive in the global AI race. I can't wait to share it with you.

2:20Let's jump in. I'm Rana El-Khalyubi and this is Pioneers of AI a podcast taking you behind the scenes of the AI revolution

2:41Please give a warm hand to Andrew Ng

2:46Alright Thank you Jay Good to see you So I was told there was a lot of heart in this audience. Yeah, there's heart in this audience. And now we go to AI because none of one's heard any of this, right, about AI. But here's the important thing. Here's the really important thing. I've known Andrew for a long time. He's been talking and working on AI long before it was cool. I remember actually sitting down with Andrew when we were coming up with these ideas around data science and these things. and he was talking about AI. I was like, hey, we're still in the middle of winter for AI. But he's been on this for the incredible journey.

3:24But to just give you some of the highlights of Andrew, he was one of the first people to advocate in using DPUs for deep learning. He wrote... Should have bought NVIDIA stock. I was going to say, how many jets does Jensen owe you? I don't know. I mean, it's like what size, right? Happy for the guys. Exactly. Did he ever give you anything? I don't think. He gave me a few GPUs. That was nice. A few GPUs. You got some GPUs out of it. All right. You wrote the first major online course on machine learning and AI, which led then to Coursera. At the time, it was radical thinking to actually teach an online course.

4:03And it's helped over 10 million students. Is that correct? 10 million? Yeah. Yeah. Thank you. Incredible accomplishment. along with Jeff Dean, Greg Corrado, and Rajat Manga. You started the Google Deep Mind, not the Deep. Google Brain. Google Brain, thank you. The Google Brain Project. Much, much more. But some of the things that you're doing right now, you have a fund and a studio investing, building on AI. And you also have some of the most seminal cited papers in AI today. I see. I have to admit, I probably have highly cited papers. I don't track my citation column as often as I used to. Exactly.

4:46You leave that for everybody else or the AI systems. But we're also, something that you don't know about, how many people have heard about agentic systems these days? Yeah, maybe I should ask it the other way around, just to highlight people. Guess where agentic came from? This is the guy. It's a little known fact that Andrew was actually the guy who really came up with agentic. Actually, let's start with that. what's the story behind Agentic? So like almost two years ago, I saw this rising trend in AI that a lot of people were excited about. But within the tech community, there was all this debate where some people would write software and say it's an agent.

5:22Others say, no, that's not an agent. It's an agent, not an agent. I thought, this is a waste of time. Why don't we, instead of having a binary, agent's not agent, let's just call it all Agentic and then stop arguing and get on with the work. And so I actually kind of ran a campaign that I didn't publicize, but I did it anyway to try to get more people to just adopt the word agentic. What I didn't realize was a few months later, a bunch of marketers would get hold of this word and slap it as a sticker on everything in sight. And that helped the movement take off. But even though the hype's gone like that, I think the real value is really growing rapidly too.

5:56So that's been exciting. Well, let's stay on that. So I want to do this in four stages. Let's first talk about today. And talking about today in the state of AI through the lens of one of the OGs, What is your honest take of what AI can and can't do, especially through this lens of agentic? Because I think we're all struggling with all the marketing and buzz out there of what's real and what's not. A lot of work that lies ahead is to take this amazing agentic AI capabilities and map it to real business workflows. I think we've been doing that for like a year. What is an agentic? Like, let's ground us there.

6:33So a lot of us use AI, large language models, by prompting it and asking it to write an output, that's a bit like going to a human on discussing AI and saying, please write an essay by just typing it out from the first word to the last word all in one go without stopping to think or without ever using backspace. So humans don't do our best writing like that and neither does AI. With a genetic workflow, the idea is we can ask an AI to take a more iterative approach and say, first write an outline, then do some work research, then write a first draft, then critique it. And so the iterative workflow takes much longer, but for a lot of tasks from medical advice, legal advice, tariff compliance, writing code for a lot of different things, this agentic workflows work much better.

7:13But there's still a lot of work ahead of us. I know that some people say, oh, don't worry about that, wait for AGI, that'll solve all the problems. I'm not a fan of this, let's wait for AGI, you know, that feels hypey to me. A lot of the work that's very valuable that you're doing today is to take the technology and what may be possible in the next six to 12 months and just go do valuable stuff with it. and and what do you what's where where sometimes the analogy i think of is as i'm using these systems because i'm trying to build uh you know i'm trying to deploy ai to really help senior citizens in their health care journey different things and sometimes i think am i on thick ice am i on thin ice with what the systems can and can't do i suspect a lot of people out there are wondering, does it work for this problem or is it fragile?

8:02Sometimes we just get to an 80 % solution. Other times it knocks out of the park. Sometimes we're incredibly disappointed. How do you as Andrew, who's building companies, advising so many people, think about this? Yeah, it is tough. I think the closer it tosses to only text processing and if you have the plumbing to get all the information, hopefully in text that people need to do a task, the easier it is for AI to do it. if you need to feed in images, voice conversations, it's not impossible, but it gets harder. And then one question I often ask is, humans know a lot of stuff, which has a lot of context.

8:36And so do we have the data plumbing to get the AI system similar context as a human would need to do that task? And then I think for a lot of multi-step processes, if you can write a standard operating procedure, like an SOP, that also may be assigned as we're seeing if we can codify the SOP in a multi-step agentic workflow. So it's hard to determine what can and cannot be done. But I think these are maybe some suggestions for rating what's more or less likely to succeed. We'll be back with more from DJ and Andrew after a quick break. Thank you.

9:38face of environmental activism, how to succeed at failing, and whether public transportation should be free. Go ahead, listen to Freakonomics Radio wherever you get your podcasts.

9:54Let's switch gears and go to education, because you've really changed and transformed academia through Coursera. Your current online courses have, I can't even keep track of how many views and how many people are taking your courses. And I want to start with, I'm sure every parent is asking you what their kid should do to be prepared for AI. Should kids still learn to code? Is CS a thing? Is data science a thing? Or is that a bad idea? Yeah, so one of the most important skills for the future is the ability to tell a computer exactly what you want so they can do it for you. and for the foreseeable future, people that know the language of computers, people that understand coding, will really do that much more effectively than people that don't.

10:41So I know that even, you know, earlier this year, there were some leaders that were advising others not to learn to code on the grounds AI will automate it. I think we'll look back on that as some of the worst career advice ever given. I'm already seeing on my teams, a lot of Silicon Valley teams, not just the software engineers, but the marketers, HR professionals, the analysts, the finance professionals, the ones that know how to code, they're starting to run circles around the ones that don't. So if your kid intends to be a software engineer, have them learn to code with AI. And even if they don't, it is becoming clearer that in the future, I think we need a lot more, not just users of software, but creators of software.

11:21And so rather than your kid growing up and asking, is there an app for that? I wanted to say, I built an app for that. And with AI assistance, coding is much easier than it used to. So don't code by hand, get AI to do it for you and people that do that will be more powerful and more effective than people that don't. So that's one challenge that the digital system has to go through as well. There's this new skill. Just like today, I can't imagine, you know, if you go through college without learning how to do web search, right? You know, that's kind of weird. It limits your job prospects. In the future, I think if you go through college and come out not knowing how to create software, we'll go, oh, that's kind of weird.

11:54It will limit the prospects of what someone can do. When we think about when should a kid get access to this technology, what does that look like and specifically through the lens of some of the lessons i think we've really started to struggle with of around social media surgeon general previously varek murthy really highlighted the challenges that are happening around that when is too early in your idea when is it the appropriate time for someone to really start having access to ai to make sure they're truly AI native and get the maximum benefits of this technology? I know I think it's difficult so I feel like you know when I was when should kids get access to books we think really really young but there are also some books that are clearly inappropriate for a two-year-old and I think one of the challenges of technology is I think there are apps that are just fine for a very young child to use but there are also a lot of stuff that we would not let a young child use so So my kids four and six, they do use tablets occasionally, but I'm there with them, right?

12:59When they're using it, I'm kind of not using as a babysitter, but having them do educational things or doing weird things when we talk about it. So I think the medium has changed from other things to tech, but I think the challenge is what are the business intents for companies to do or not create certain experience for the kids? And as parents, how can we have guardrails or curate the hopeful things. Just like I don't let my books because we read certain highly inappropriate books for their age, I don't let them do certain highly inappropriate things for their age. But it's a challenge, giving the incentives of certain types of companies to do things that we as parents may not want them to.

13:38Do you ever talk to some of those companies or groups and you're like, hey, knock it off? That's not going to be helpful? Or what's that conversation like? Because these are a lot of people who have taken your classes that are actually doing some of the behaviors that I think as parents, we find really problematic. You know, 99 % of engineers and business people in Silicon Valley want to do the right thing, right? The people doing these, frankly, they're, you know, our friends, maybe some people in this room right now. I think everyone kind of wants to do the right thing. And I wish we could find a way when the billions of dollars at stake to still always do the right thing.

14:15It is a real problem. And we do see a small number of people that will sometimes do not quite the right thing when the financial incentives or some other incentives are big enough. I wish I knew how to solve the problem of human incentives, I guess. Great.

14:32Well, let's switch to something easier. U.S. policy. Speaking of children.

14:45Statements that former U.S. chief data scientists get to make. You were early in advocating both, you know, as we mentioned, the policies of GPU, but also, you know, you were one of the first people I saw that really were working with international technology companies. And so given some of the place where you sit and you get to see across the landscape, you've seen, we've seen this like whipsawing on GPUs from federal policy. We've seen executive orders talking about what they would describe in their words, woke systems. We've seen also executive orders and policies trying to accelerate the adoption of AI.

15:24And so if you had five minutes with the president, what advice would you give them to both, A, responsibly unleash the power of AI to benefit all Americans, and B, ensure national competitiveness? What would that be? Yeah, I'm really worried about U.S. national competitiveness in AI. I think some things that the current administration has done well, I think the previous administration has some AI safety types of thinking that was really safety theater driven by lobbyists, fear-mongering to try to create, regulatory capture, anti-open source type of regulations, right? So if you don't want to compete with open source, make up a bunch of stuff about the dangers of AI to try to get stifling, licensing, whatever parts.

16:07So I think current administration, you know, seems to have very low patience with that. that's good. Things that worry me, I think that both of us are immigrants. I think that a lot of our students are immigrants. I really worry about American competitiveness if we make it harder for high school immigration. But frankly, not just high school immigration. When I came to the United States as an undergrad student, I think I was like 17 years old, I was frankly pretty clueless. I don't think I was high school at all when the U.S. let me in. So letting students come to the US and then grow up and hopefully be higher skill, I really worry about that.

16:42I think defunding of science, the decreased investments in science and AI and other things. Universities do have issues, right? There are things we can fix at universities, but I think that diminishing the ability of this country to execute science, I really worry about that. And then I think in terms of national policy, I also worry about our reliance on TSMC, I think TSMC Arizona's chip maker. It's been interesting to see China recently ban certain imports of NVIDIA chips, which is a strong signal that China is moving toward independence from TSMC in Taiwan at a moment where the US is becoming still heavily reliant on Taiwan manufacturing.

17:29One of the implications of this is if anything were to happen in Taiwan, either a natural disaster or a man-made event, disruption to the Taiwan semiconductor ecosystem could end up hurting the US much more than it hurts China if China becomes more independent of Taiwan manufacturing than the US. So I think semiconductors, and then lastly, I think that AI semiconductor is a bottleneck. The other big bottleneck, which you read up in the news, is totally true. It is energy. But when you build a data center, that's a machine to turn electricity into intelligence or turn electricity into arm tokens.

18:05And so the constraints, I have so many friends that are hung up in permitting, can we build a power plant here? And then you think you're the permits, but then there's your local objections or whatever, which are maybe valid. But I think that the energy capacity is the other bottleneck that I really worry about. When you think about China specifically, let's just take China and Europe. So two very radically different approaches in how they think about AI, regulation, owning the thing. And then you watch the U.S. trying to navigate this for competitiveness. One of the things that I'm curious about from your lens is this idea of both the open source models and do people build off of the main trunk of AI, which is U.S.-based primarily right now, but increasingly new models out of China, a little bit out of Europe.

19:01What's the right strategy here? What's a strategy that you believe is best and optimal for the world for AI? Is it kind of like a central trunk of AI with branches? Or is it many different trees in the forest with people with a federation of different models, approaches, techniques? I think we need multiple branches. Because otherwise, one of the reasons why, I make an analogy, mobile phones, the mobile ecosystem is kind of uninteresting. It's because there are two gatekeepers, Android and iOS. and unless they let you do certain things, you're just not allowed to experiment. So I hope the AI will not end up with a small number of gatekeepers that can limit innovation.

19:41What has happened is over the past year, two years, China has really pulled ahead of the US in releasing open-weight models. These are models anyone in the world can download and use for free. And I think I saw a stat showing that cumulative adoption of Chinese open-weight models I think is about to surpass or may already have surpassed cumulative adoption of US open-weight models. The US closed models are still better, but open-weight models are a key part of the AI supply chain and people are using them. I think there's a problem that we aren't investing enough as a nation in that. And then you asked about Europe.

20:16Honestly, I think, love Europe. I wish Europe would wake up and get going faster. For a while, over the last few years, visiting European regulators, I heard things like, We want to be leaders in regulating AI. You know, I don't think that's how you gain competitive advantage. More brakes, more brakes, less gas. Win the race. More in a minute. Stay with us.

20:51Meet Nicole Nicholas, Capital One business customer and co-owner of Ansett Uncles, a plant-based restaurant and community space in Brooklyn, New York, that got its start from a need for unity. The inspiration, it was born from the desire to create a space that felt like home, where we can connect community culture, good food, and come together with family and friends. That's how we birthed aunts and uncles. Nicole and her husband, Mike, were fulfilling their dream of bringing people together out of their home kitchen. But they soon learned that the demand for community was greater than they knew.

21:22It became overwhelming and we were like, we need home, but not in our actual home. We realized that there was also a need in our community for something bigger in our neighborhood. So we had to find a place. Moving from a home operation into a storefront was a huge next step. But Nicole and Mike were able to take it on with the help of Capital One Business. It's not for the weak. As a small business, finding resources is super important because that's the way you'll be able to manage and scale. We would have never done that without having Capital One to be able to help us along the way. The cashback rewards are very helpful.

21:58You know, it just gave us that runway to be able to breathe a little bit. Then you get to focus on the cooking of the food and making the experience great. To learn more, go to CapitalOne.com slash business cards.

22:14I want to turn to the future in this remaining couple minutes and really talk about the future of AI. but I want to talk about it through the cutting edge next generation of students that you see, the entrepreneurs. What are the problems that you see them gravitating to? What are the things, the hopes, the dreams? And specifically, what does that tell you about how the next 24 months look? So we're in Silicon Valley where most of us love AI. I love AI, love what I do, I think it makes the world better. I think many of us may underestimate the distrust that a lot of people across the nation have for AI.

22:57And I feel urgency to kind of get our act together to make sure that we can tell a compelling narrative to explain why AI is actually good for the world. It turns out we're all getting excited about productivity improvements, but when a contact center worker is scared of losing their job, We're a fast food worker. Here's a politician saying, yep, guess what? These AI people, they're going to make your job go away. That creates a lot of fear and distrust of AI that we don't really see here in Silicon Valley. So I think to win people over, we need to make sure technology genuinely benefits everyone very large.

23:36And I think there is a path to that. AI can make individuals much more effective and much more productive. But to get the tools available to everyone, teach everyone to use it, that's the upskilling, improve the tools. And really, you know, I feel like with the concept of a 10x engineer, I think with AI, we can have 10x marketers, 10x analysts, 10x finance professionals. But to actually make that happen, it feels like there's a lot of work ahead of us. And I worry that we have not yet won the trust of a lot of people in this country. What's your favorite way that you use AI today? Gosh, maybe I'll share one that is not widely known.