Kevin Scott: Microsoft CTO | Lex Fridman Podcast #30

The following is a conversation with Kevin Scott, the CTO of Microsoft. Before that he was the Senior Vice President of Engineering and operations at LinkedIn and before that he oversaw mobile ads engineering at Google. He also has a podcast called behind the Tech with Kevin Scott, which I'm a fan of. This was a fun and wide ranging conversation that covered many aspects of computing. It happened over a month ago before the announcement of Microsoft's investment OpenAI that a few people have asked me about. I'm sure there'll be one or two people in the future that'll talk with me about the impact of that investment. This is the Artificial Intelligence podcast. If you enjoy it, subscribe on YouTube, give it five stars on iTunes, support it on Patreon, or simply connect with me on Twitter exfriedman spelled F R I D M A N and I'd like to give a special thank you to Tom and Alanti Bickhousen for their support of the podcast on Patreon. Thanks Tom and AlanThe hope I didn't mess up your last name too bad. Your support means a lot and inspires me to keep this series going. And now here's my conversation with Kevin Scott.

You've described yourself as a kid in a candy store at Microsoft because of all the interesting projects that are going on, can try to do the impossible task and give a brief whirlwind view of all the spaces that Microsoft is working in, both research and product.

If you include research, it becomes even more difficult. So I think broadly speaking, Microsoft's product portfolio includes everything from big cloud business, like a big set of SaaS services. We have the original or some of what are among the original productivity software products that everybody uses. We have an operating system business, we have a hardware business where we make everything from computer mice and headphones to high end personal computers and laptops. We have a fairly broad ranging research group where we have people doing everything from economics research. So there's this really smart young economist, Glenn Weil, who like my group works with a lot, who's doing this research on these things called radical markets. Like he's written an entire entire technical book about this whole notion of radical markets. So like the research group sort of spans from that to human computer interaction to artificial intelligence. And we have GitHub, we have LinkedIn, we have a search, advertising and news business and like probably a bunch of stuff that I'm embarrassingly not recounting in this list.

Gaming to Xbox and so on, right?

Yeah, gaming for sure. Like I was having a super fun Conversation this morning with Phil Spencer. So when I was in college, there was this game that LucasArts made called Day of the Tentacle that my friends and I played forever. And we're doing some interesting collaboration now with the folks who made Day of the Tentacle. And I was completely nerding out with Tim Schafer, like the guy who wrote Day of the Tentacle this morning. Just a complete fanboy, which, you know, sort of, it happens a lot. Like Microsoft has been doing so much stuff at such breadth for such a long period of time that being cto, most of the time my job is very, very serious. And sometimes I get caught up in how amazing it is to be able to have the conversations that I have with the people I get to have them with.

You have to reach back into the sentimental and what's the radical markets and the economics.

So the idea with radical markets is can you come up with new market based mechanisms to. I think we're having this debate right now. Does capitalism work, free markets work? Can the incentive structures that are built into these systems produce outcomes that are creating sort of equitably distributed benefits for every member of society? And I think it's a reasonable set of questions to be asking. And so what, Glenn? And so one mode of thought there, if you have doubts that the markets are actually working, you can sort of tip towards like, okay, let's become more socialist and have central plans and governments or some other central organization is making a bunch of decisions about how work gets done and where the investments and where the outputs of those investments get distributed. Glenn's notion is lean more into the market based mechanism. So for instance, this is one of the more radical ideas. Suppose, suppose that you had a radical pricing mechanism for assets like real estate where you were, you could be bid out of your position in your home, for instance. So like if somebody came along and said, I can find higher economic utility for this piece of real estate that you're running your business in, like, then you either have to bid to stay or the thing that's got the higher economic utility takes over the asset, which would make it very difficult to have the same sort of rent seeking behaviors that you've got right now. Because if you did speculative bidding, like you would you very quickly like lose a whole lot of money. And so like, the prices of the assets would be sort of like very closely indexed to like the value that they could produce. And like, because like you'd have this sort of real time mechanism that would force you to sort of mark the value of the asset to the market, then it could be taxed appropriately. Like, you couldn't sort of sit on this thing and say, oh, like this house is only worth 10,000 bucks when everything around it is worth 10 million.

That's finished. So it's an incentive structure where the prices match the value much better.

And Glenn does a much better job than I do at selling. And I probably picked the world's worst example. And it's intentionally provocative. So this whole notion, I'm not sure whether I like this notion that we could have a set of market mechanisms where I could get bid out of my profit property. But if you're thinking about something like Elizabeth Warren's wealth tax, for instance, you would have. I mean, it'd be really interesting in how you would actually set the price on the assets. And you might have to have a mechanism like that if you put a tax like that in place.

It's really interesting that that kind of research, at least tangentially, is touching Microsoft research.

Yeah.

That you're really thinking broadly. Maybe you can speak to. This connects to AI. So we have a candidate, Andrew Yang, who kind of talks about artificial intelligence and the concern that people have about automation's impact on society. And arguably, Microsoft is at the cutting edge of innovation in all these kinds of ways, and so is pushing AI forward. How do you think about combining all our conversations together here with radical markets and socialism and innovation in AI that Microsoft is doing, and then Andrew Yang's worry that will result in job loss for the lower and so on? How do you think about that?

I think it's sort of one of the most important questions in technology, like, maybe even in society right now, about how is AI going to develop over the course of the next several decades and what's it going to be used for? And like, what. What benefits will it produce and what negative impacts will it produce? And, you know, how. Who. Who gets to steer this whole thing? You know, I'll say it at the highest level. One of the real joys of getting to do what I do at Microsoft is Microsoft has this heritage as a platform company. And so, you know, like Bill. Bill has this thing that he said a bunch of years ago where the measure of a successful platform is that it produces far more economic value for the people who build on top of the platform than is created for the platform owner or builder. And I think we have to think about AI that way. Yeah, it has to be a platform that other people can use to build businesses to fulfill their creative objectives, to be Entrepreneurs to solve problems that they have in their work and in their lives. It can't be a thing where there are a handful of companies sitting in a very small handful of cities geographically, who are making all the decisions about what goes into the AI. And then on top of all this infrastructure, then build all of the commercially valuable uses for it. So I think that's bad from a sort of economics and equitable distribution of value perspective back to this whole notion of do the markets work? But I think it's also bad from an innovation perspective because I have infinite amounts of faith in human beings that if you give folks powerful tools, they will go do interesting things. And it's more than just a few tens of thousands of people with the interesting tools. It should be millions of people with the tools. So it's sort of like you think about the steam engine in the late 18th century. It was maybe the first large scale substitute for human labor that we've built like a machine. And in the beginning, when these things are getting deployed, the folks who got most of the value from the steam engines were the folks who had capital so they could afford to build them. And like they built factories around them and businesses and the experts who knew how to build and maintain them. But access to that technology democratized over time. Like now, like, like an engine is not a. It's not like a differentiated thing. Like there isn't one engine company that builds all the engines and all of the things that use engines are made by this company. And like they get all the economics from of that. No, no, like fully Democrat. Like they're probably, you know, we're sitting here in this room and like, even though they don't, they're probably things, you know, like the, the MEMS gyroscope that are in both of our phones. Like they're like little engines, you know, sort of everywhere. They're just a component in how we build the modern world. Like AI needs to get there.

Yeah.

So that's a really powerful way to think. If we think of AI as a platform versus a tool that Microsoft owns as a platform that enables creation on.

Top of it, that's a way to democratize it.

That's really interesting actually. And Microsoft throughout its history has been positioned well to do that.

And the tie back to this radical markets thing, my team has been working with Glenn on this and Jaron Lanier, actually. So Jaron is the, like the sort of father of virtual reality. Like, he's one of the most interesting human beings on the planet. Like a sweet Sweet guy. And so Jaron and Glenn and folks in my team have been working on this notion of data is labor or like, they call it data dignity as well. And so the, the idea is that if you, again, going back to this sort of industrial analogy, if you think about data as the raw material that is consumed by the machine of AI in order to do useful things, then we're not doing a really great job right now in having transparent marketplaces for valuing those data contributions. So like, and we all make them like explicitly. Like you go to LinkedIn, you sort of set up your profile on LinkedIn, like that's an explicit contribution. Like you know exactly the information that you're putting into the system. And like, you put it there because you have some nominal notion of like, what value you're going to get in return. But it's like only nominal. Like you don't know exactly what value you're getting in return. Like service is free, you know, like it's low amount of like perceived. And then you've got all this indirect contribution that you're making just by virtue of interacting with all of the technology that's in your daily life. And so like, what Glenn and Jaron and this data dignity team are trying to do is like, can we figure out a set of mechanisms that let us value those data contributions so that you could create an economy and like a set of controls and incentives that would allow people to, like, maybe even in the limit, like, earn part of their living through the data that they're creating. And like, you can sort of see it in explicit ways. These companies, like scale AI and like there are a whole bunch of them in China right now that are basically data labeling companies. So like you're doing supervised machine learning, you need lots and lots of label training data. And like those people are getting compute work for those companies are getting compensated for their data contributions into the system.

And so that's easier to put a number on their contribution because they're explicitly labeling data. Correct. But you're saying that we're all contributing data in different kinds of ways. And it's fascinating to start to explicitly try to put a number on it. Do you think that's, that's possible?

I don't know. It's hard. It really is because, you know, we don't have as much transparency as is. I think we need in like how the data is getting used and it's, you know, super complicated. Like, you know, we, we, I think as technologists sort of appreciate some of the subtlety There it's like the data gets created and then it gets. It's not valuable like the data exhaust that you give off or the explicit data that I am putting into the system isn't valuable, isn't super valuable atomically. It's only valuable when you sort of aggregate it together into sort of large numbers. It's true even for these folks who are getting compensated for labeling things for supervised machine learning now you need lots of labels to train a model that performs well. And so I think that's one of the challenges. It's like, how do you sort of figure out, because this data is getting combined in so many ways through these combinations, how the value is flowing?

Yeah, that's fascinating. Yeah, and it's fascinating that you're thinking about this. And I wasn't even going into this conversation expecting the breadth of research really that Microsoft broadly is thinking about. You are thinking about at Microsoft. So if we go back to 89, when Microsoft released Office or 1990, when they released Windows 3.0, how's the, in your view? I know you weren't there the entire, you know, through its history, but how has the company changed in the 30 years since? As you look at it now, the.

Good thing is it's started off as a platform company. Like, it's still a platform company. Like the parts of the business that are like thriving and most successful are those that are building platforms. Like the mission of the company now is the missions change. It's like change in a very interesting way. So back in 89, 90, they were still on the original mission, which was put a PC on every desk and in every home. And it was basically about democratizing access to this new personal computing technology, which when Bill started the company, integrated circuit microprocessors were a brand new thing. And like people were building homebrew computers from kits, like the way people build ham radios right now. And I think this is sort of the interesting thing for folks who build platforms in general. Bill saw the opportunity there and what personal computers could do. And it was like a, it was sort of a reach. Like you just sort of imagine like where things were, you know, when they started the company versus where things are now. Like in success. When you've democratized a platform, it just sort of vanishes into the platform. You don't pay attention to it anymore. Like operating systems aren't a thing anymore. Like they're super important, like completely critical. And like, you know, when you see one fail, like you just, you sort of understand but like, you know, it's not a thing where you're not like waiting for the next operating system thing in the same way that you were in 1995. Right. Like in 1995, like, you know, we had rolling stuff on the stage with the Windows 95 rollout. Like, it was like the biggest thing in the world. Everybody lined up for it the way that people used to line up for iPhone. But eventually, and this isn't necessarily a bad thing, it just sort of, the success is that it's sort of. It becomes ubiquitous. It's like everywhere. And like human beings, when their technology becomes ubiquitous, they just sort of start taking it for granted. So the mission now that Satya re articulated five plus years ago now, when he took over as CEO of the company, our mission is to empower every individual and every organization in the world to be more successful. And so again, that's a platform mission. And the way that we do it now is different. It's like we have a hyperscale cloud that people are building their applications on top of. Like, we have a bunch of AI infrastructure that people are building their AI applications on top of. We have a productivity suite of software like Microsoft Dynamics, which some people might not think is the sexiest thing in the world, but it's like helping people figure out how to automate all of their business processes and workflows and to help those businesses using it to like, grow and be more so. It's, it's a much broader vision in a way now than it was back then. Like, it was sort of very particular thing. And like now, like, we live in this world where technology is so powerful and it's like such a basic fact of life that it, you know, that it both exists and is going to get better and better over time, or at least more and more powerful over time. So, like, you know, what you have to do as a platform player is just much bigger.

Right.

There's so many directions in which you can transform. You didn't mention mixed reality too. You know, that's. Yep, that's. That's probably early days or it depends how you think of it. But if we think on a scale of centuries, it's the early days of mixed reality.

Oh, for sure.

And so, yeah, with HoloLens, Microsoft is doing some really interesting work there. Do you touch that part of the effort? What's the thinking? Do you think of mixed reality as a platform too?

Oh, sure. When we look at what the platforms of the future could be, it's fairly obvious that AI is one you don't have to. You sort of say it to someone and they get it. But we also think of the mixed reality and quantum. Is these two interesting potentially quantum computing? Yeah.

Okay, so let's get crazy then. So you're talking about some futuristic things here. Well, the mixed reality Microsoft is really not even futuristic. It's here it is incredible stuff.

And look and it's having impact right now. Like one of the more interesting things that's happened with mixed reality over the past couple of years that I didn't clearly see is that it's become the computing device for folks for doing their work who haven't used any computing device at all to do their work before. So technicians and service folks and people who are doing machine maintenance on factory floors so they, because they're mobile and they're out in the world and they're working with their hands and sort of servicing these very complicated things. They don't use their mobile phone and they don't carry a laptop with them and they're not tethered to a desk. And so mixed reality where it's getting traction right now, where hololens is selling a lot of units is for these sorts of applications for these workers. And it's become like, I mean like the people love it. They're like oh my God, like this is for them like the same sort of productivity boost that an office worker had when they got their first personal computer.

Yeah, but you did mention it's fairly obvious AI as a platform. But can we dig into it a little bit? How does AI begin to infuse some of the products in Microsoft? So currently providing training of for example neural networks in the cloud or providing pre trained models or just even providing computing resources, whatever different inference that you want to do using neural networks. How do you think of AI infusing as a platform that Microsoft can provide?

Yeah, I mean I think it's, it's super interesting. It's like everywhere and like we, we run these, we run these review meetings now where it's me and Satya and like members of Satya's leadership team and like a cross functional group of folks across the entire company who are working on like either AI infrastructure or like have some substantial part of their, of their product work using AI in some significant way. Now the important thing to understand is like when you think about how the AI is going to manifest in an experience for something that's going to make it better, I think you don't want the AI ness to be the first order thing. It's like whatever the product is and the thing that is trying to help you do, the AI just sort of makes it better. And this is a gross exaggeration, but people get super excited about where the AI is showing up in products. And I'm like, do you get that excited about where you're using a hash table in your code? It's just another.

It's just a tool.

It's a very interesting programming tool, but it's sort of an engineering tool and so it shows up everywhere. So we've got dozens and dozens of features now in Office that are powered by fairly sophisticated machine learning. Our search engine wouldn't work at all if you took the machine learning out of it. Increasingly, things like content moderation on our Xbox and xCloud platform.

When you mean moderation, do you mean the recommenders is showing what you want to look at next?

No, it's like anti bullying stuff.

So the usual social network stuff that you have to deal with.

Yeah, correct. But it's like really, it's targeted, it's targeted towards a gaming audience. So it's like a very particular type of thing where, you know, the, the line between playful banter and like legitimate bullying is like a subtle one and like, you have to like, it's sort of tough. Like I have.

I'd love to if we could dig into it because you're also. You led the engineering efforts of LinkedIn. Yep. And if we look at, if we look at LinkedIn as a social network.

Yep.

And if we look at the Xbox gaming as the social components, the very different kinds of, I imagine communication going on on the two platforms. Right. And the line in terms of bullying and so on is different on the platforms. So how do you, I mean, it's such a fascinating philosophical discussion of where that line is. I don't think anyone knows the right answer. Twitter folks are under fire now, Jack at Twitter for trying to find that line. Nobody knows what that line is, but how do you try to find the line for, you know, trying to prevent abusive behavior and at the same time let people be playful and joke around and that kind of thing?

I think in a certain way, like, you know, if you have what I would call vertical social networks, it gets to be a little bit easier. So if you have a clear notion of what your social network should be used for or what you are designing a community around, then you don't have as many dimensions to your content safety problem as you do in a general purpose platform. So on, on LinkedIn, the whole social network is about connecting people with opportunity, whether it's helping them find a job or to find mentors or to help them find their next sales lead, or to just allow them to broadcast their professional identity to their network of peers and collaborative.