No Priors 🎙️119: Going Full Send on AI and the (Positive) Impact of AI on Jobs, with Kevin Scott, CTO Microsoft (TRANSCRIPT)
In this episode, Sarah and Elad speak with Microsoft CTO Kevin Scott about his unlikely journey from rural Virginia to becoming the driving force behind Microsoft's AI strategy.
Sarah and Elad discuss the partnership that Kevin helped forge between Microsoft and OpenAI and explore the vision both companies have for the future of AI. They also discuss yesterday’s announcement of “copilots” across the Microsoft product suite, Microsoft’s GPU computing budget, the potential impact of open source AI models in the tech industry, the future of AI in relation to jobs, why Kevin is bullish on creative and physical work, and predictions for progress in AI this year.
No Priors is now on YouTube! Subscribe to the channel on YouTube and like this episode.
- May 23, 2023: The Verge - Microsoft CTO Kevin Scott Thinks Sydney Might Make a Comeback
- May 23, 2023: Microsoft Outlines Framework For Building AI Apps and Copilots
- January 10, 2023: A Conversation with Kevin Scott: What’s Next In AI
[00:00] - Kevin Scott's Journey to Microsoft CTO
[12:44] - Microsoft and Open AI Partnership
[21:18] - The Future of Open Source AI
[32:12] - AI for Everyone
[45:29] - AI and the Future of Jobs
[51:44] - The Future of AI and Regulation
[58:10] - Taking a Global Perspective
Sarah Guo: Microsoft, the BMF productivity cloud and gaming company, has taken a massive bet on AI. Everyone's paying close attention to its partnership with OpenAI and the technical community has been amazed by its release of some of the first truly useful and broadly deployed AI products, such as GitHub Copilot. It's full on attack on web search with the new LLM powered Bing Chat is making its incumbent competitors dance. Today on No Priors we're thrilled to speak with Kevin Scott, CTO of Microsoft and the driving force behind their AI strategy. Kevin's leadership, both at Microsoft and prior at LinkedIn, Google, and AdMob as a technologist is especially inspiring to me, given his distance traveled from his childhood home in rural Central Virginia. In 2020, he published a book, Reprogramming the American Dream, about making AI service all. Kevin, welcome to No Priors. Thanks so much for joining us.
Kevin Scott: Thanks for having me, guys.
Sarah Guo: Can you start by sharing with us some of your story? How does one go from a farming community in Virginia, where your parents didn't attend college, to CTO of Microsoft?
Kevin Scott: I don't know. I think it is a very unlikely journey, certainly not a thing that I ever could have imagined. I think part of it is I was just super fortunate to be wired like a nerd and growing up when I grew up. When I was a teenager in the early '80s, personal computing was happening and that was the thing that I happened to fixate on. And even though we were relatively poor, I managed to scrape together enough bucks to get myself a personal computer that I could have and just tinker with all the time. It was a Radio Shack Color Computer 2, one of these things with Chicklet keys that you actually connected to a television. I had it hooked up to a 13-inch TV and it had a cassette recorder that you stored and loaded your programs on.
It was just the thing that I was obsessed with and I stayed obsessed with computers from then on and it was just me trying to find a path at each step where I could work on the most interesting thing that someone was dumb enough to give me permission to go work on. Again, it's a lot of luck. There's no way you can plan a path from rural Central Virginia to CTO of Microsoft, but I think it does help to have a high level vision in your head for what it is that you want to do. Just knowing what you're aiming for always helps.
Sarah Guo: What was that vision for you besides obsessed with computers, wanted to work on them?
Kevin Scott: Yeah. I more or less had two of them. The first vision I had when I was a teenager was I wanted to be a computer science professor, so I just looked at what computer scientists did and thought, "This is the most amazing stuff I've ever seen." I went to a science and technology high school, and the way that it worked where I lived is a really rural area, and so the science and technology was a governor's school, so it was centrally located. Each high school in these four or five counties that surrounded the governor's school got to send two students each, and so I was one of the two students that got selected from my high school to go to this thing.
My computer science professor there was this guy, Dr. Tom Morgan, and I just felt like he'd opened up this entire new world to me. It was just thrilling to learn all of this stuff and I was like, "Yeah, I want to be like Dr. Morgan." A lot of this stuff for me is about who those influential role models have been in your life, and so as soon as I met Dr. Morgan, I was like, "Oh, I should just go be a computer science professor." That was the path I was on until I was about 30 years old, when I was a compiler optimization and computer architecture programming languages person, and I got pretty disillusioned with what being a computer science professor actually was relative to what I wanted to do.
I just wanted to have a lot of impact and my perception at the time when I was making these decisions was that you could have a lot of impact as a computer science professor and the impact was actually great, but it wasn't the impact that the system appreciated. The impact that you can actually have is inspire students to go pursue these careers and they will go on to do much greater things than you've done yourself. And that, to me, was the greatest impact, but it was the least appreciated part of being a computer science professor back in the 2000s, when I was making these big decisions, and so I decided to leave and I didn't at the time know what next actually was going to be. It had been my mission for almost 15 years at that point and I was a little bit lost and I saw that a bunch of my academic buddies were all working at this startup called Google and I didn't understand why they were working at Google. Google was some little box and you typed a keywords in and it gave you 10 links. How is that hard?
Urs Holzle, who was a compiler person, and Jeff Dean, who was a compiler person, and Alan Eustace, who was a compiler person, all of these people who I went to conferences with and whose papers I read and I was like, "All right, well, maybe I should send my resume in." I sent my resume in and got called to do a bunch of interviews. It was the best interviewing experience I've ever had because they took what must have been every compiler person in the company at the time and put them on my interview panel and I was like, "Oh my God, this is amazing." I had the best day interviewing there and I got this job offer and I got this choice. They had just started Google New York, which was the first office outside of Mountain View, and they were like, "You can come to Mountain View or you can go be the 10th person in this New York office." My wife and I wanted to live in New York more than we wanted to live in Mountain View, and so that's what we did.
After I got there, this is where the new mission came in. We were hiring these brilliant, brilliant people at the time, and the way that we did hiring was crazy. It's like, "All right. Well, if you're smart, just come work here and we have no idea what exactly it is you're going to do," and you came in and you sorted yourself out. We had these people who were so accomplished and so brilliant and they would come in and choose to work on things that just were going to have no impact at all. They were intellectually very interesting, but they were just silly, in that they were never going to connect with anything that moved the needle for the company, which was exactly the problem I was trying to get away from in being a research computer scientist, and so I sorted myself out. I found a pragmatic thing to go work on.
I won't go into the details of what it is, but the whole team won a Google Founder's award, which was a big deal for solving this very unsexy problem with a bunch of very fancy computer science, which was one of the things I think Google did really well. Then I was like, "Okay. Well, I should just go help more people sort themselves out as well," and that's when I became a manager. Then from that point on, it was all about, "Hey, I want to help as many engineers as I possibly can, make sure that their work lines up with something that's both interesting and meaningful."
Elad Gil: I think that it's actually pretty underdiscussed, to a degree, to which early Google had so many academics actually running important parts of the company. I think Urs is a great example and I think there's others. I haven't actually seen anything like that since maybe now more recently at OpenAI, there's more academics or you feel like the research community's popping back up again, but it's been maybe a decade or two since that's happened.
Kevin Scott: Yeah. I mean, I think that's actually a really, really great observation. When I go sit in OpenAI, it really reminds me of early Google days and it's about the same size Google was when I joined, and so I couldn't figure it out for a while and I was like, "Wow, this is really giving me early Google nostalgia." The conclusion to draw from that is not that they're the same companies or they're trying to solve the same problem. It's just the energy of the place and who they've chosen to hire.
Elad Gil: Yeah. It's the first time I've seen string theorists getting hired again into computer science roles-
Kevin Scott: Yeah, a hundred percent.
Elad Gil: ... since Google days. Yeah.
Kevin Scott: You and I probably both worked with Yonathan Zunger, who works at Microsoft right now. I remember I was like, "All right. Yonathan's working on this big distributed file system stuff and what's his degree? Oh yeah, he's a string theory guy."
Sarah Guo: A big part of your mission for the last decades has been helping string theorists and other engineers figure out how to be useful in their orgs. The other part seems to be, of course, actual technical direction, deciding what's worth investing in, and you've worked on machine learning products for a really long time, ads auctions at Google, recommendations at LinkedIn, et cetera, et cetera. Was there a moment when you decided or you realized personally that AI should be a key technical bet for Microsoft?
Kevin Scott: Yeah. I mean, I've been at Microsoft a little over six years now, so almost six and a half years, and pretty quickly it was obvious that AI was going to be very, very, very important to the future of the company. I think Microsoft already understood that before I got there and then it was just, how do you focus all of the energy on the company on the right thing, because we had a lot of AI investment and a lot of AI energy and it was very diffuse when I got there. No lack of IQ and actually no lack of capital spending and everything else, but it was just getting peanut buttered across a whole bunch of stuff. The thing that really catalyzed what we were doing is, I mean, maybe this is a little bit too technical, but before I got there, the technical thing that had been happening with some of these AI systems that, to me, was very interesting is transfer learning was starting to work.
You were going from this mode of the flavor of statistical machine learning that I cut my teeth on in my first projects at Google, which was, you have a particular domain of data and you have a particular machine learning model architecture that you're training and then a particular way that you're going to go do the deployment and measurement and whatnot and it's all siloed to a use case or a domain or an application, to seeing AI systems that you could train on one set of data and use for multiple purposes.
You saw a little bit of that with some of the cool stuff that DeepMind was doing with reinforcement learning with play transfer across some of the gaming applications that they were building, but the really exciting thing was when it started working for language with ELMo and then BERT and then RoBERTa and Turing and a bunch of things that we were doing. That was the point where there were so many language-based applications that you could imagine building on top of these things if it continued to get better and better, and so we were just looking for evidence that it was going to continue to get better and better and as soon as we found it, we just started all in. That was everything from doing a partnership with OpenAI to, at one point I seized the entire GPU budgets for the whole company, and I was like, "We will no longer peanut butter these resources around. We will focus them because it's all capital intensive." We will just allocate these things to things where we have really, really strong evidence-based conviction that a particular path is going to benefit from adding more capital scale.
Sarah Guo: I remember it must have been five years back now, we're at dinner and now GPU capacity is the talk of the technical town, but I asked you what your most pressing issue was and you're like, "How am I going to spend on GPUs this year and how I'm going to distribute those GPUs?"
Kevin Scott: Yeah, and it was and it has been and certainly hasn't gotten any easier. I mean, Eli, I think the question you were asking is how we decided to do the open AI partnership. The reason that we did the partnership was twofold. One is with transfer learning actually working, you can imagine building a platform for all of this stuff so that you're building single things where you're amortizing the cost of the things across a whole bunch of different applications. And because we have a hyper scale cloud, one of the things that I was really, really interested in and beyond interested, it felt just an existential thing, is how do you make sure that the way that you're building your cloud all the way from your computing infrastructure, your networks, your software frameworks and whatnot, how can it really serve a whole bunch of interests beyond your own? We felt, in addition to the high ambition things that we were doing inside of the company, that we needed high ambition partners and when we looked around OpenAI was clearly the.
PART 1 OF 4 ENDS [00:14:04]
Kevin Scott: Ambition partners. And when we looked around, OpenAI was clearly the highest ambition partner that was in the field. And I think still their ambition is just breathtaking in what it is that they're trying to accomplish. And so that was one thing. And then the second thing was they really had a very similar vision to the one that I had about these things were evolving into platforms and we were able to... Because we were so aligned on vision for the future, we could figure out how to do a partnership where even though there's just a ton of difficult things. And I think there's probably some conservation law of the stress from difficulty. So it's not like it ever goes away, but it's stress in service of a common goal. And that's the thing that make good partnerships work.
Elad Gil: I think one of the stunning things about the partnership in some sense was the timing. Because if I remember correctly, Microsoft made its first investment or its first significant investment in OpenAI right after GPT-2 launched. Right around GPT-2 and this is before GPT-3, and there was such a big step function between the two of them that I think it was less obvious in the GPT-2 days that this was going to be as important as it was. And so I'm a little bit curious, what were the signs that made you decide that this was a good partnership to have versus building it internally versus... Usually, as a larger company, there's the old buy build partner kind of thinking. And so I'm just sort of curious how you all decided to partner in this moment in time where it was very non-obvious and you invested a large sum of money behind that.
Kevin Scott: Yeah. And I don't want to have revisionist history and paint a rosier picture than there actually was. So there's a huge diversity of opinions inside of the company on the wisdom of doing this. And so Satya has this thing that he talks about, no regrets investing. So they're things where you do the investment and there are multiple ways to win, and you even win a little bit when you lose. And so this was one of those no regrets things in that the very, very worst thing that could happen is we would go spend a bunch of capital on computing infrastructure and we would learn what to do at very high scale for building these AI training environments. And you'd have to believe something very strange about the world of AI that you wouldn't need advanced computing infrastructure.
And then, there were just multiple ways where, and we had a bunch of evidence that we had gathered ourselves and that OpenAI had that gave us, which unfortunately I can't talk about, but that gave us pretty reasonable confidence that scale up was actually working. You probably seen the famous OpenAI Compute Scale paper where they sort of plot on the log scale how many petaflop days or whatever the unit of total compute they were using on that graph that shows from 2012 when we first figured out how to train models with GPUs through, I think the plot ends sometime in 2018. You're basically consuming 10 times more compute every year for training state-of-the-art models. And so I just had super, super high confidence that we were never going to get to the point where we're like, all right, we got enough compute.
Elad Gil: It was a very bold move. I think it's very striking all the amazing things Microsoft has done over the last few years in terms of just incredibly smart strategic moves. Though the time didn't seem obvious and now are just in hindsight, really brilliant. I guess, a more recent move as you announced a collaboration within Nvidia to build a supercomputer part by Azure infrastructure combined with Nvidia GPUs. Could you tell us a little bit more about your supercomputing efforts in general and then maybe a little bit more about those collaborations both Nvidia and OpenAI on the supercomputing side?
Kevin Scott: Yeah, so we built the first thing that we called an AI supercomputer. I think we started working on it in 2019, and we deployed it at the end of that year. And it was the computing environment that GPT-3 was trained on. And we had been building a progressively more powerful set of these supercomputing environments. We built them in a way where the biggest environment is just because they're very capital intensive things tend to get used for one purpose. But the designs of these systems, we can build smaller stamps of them and they get used by lots of people. So we have tons of people who are training very big models on Azure compute infrastructure, both folks inside the company and partners who can come in. And it was a thing that was not possible to do before where you could say, "Hey, I would like a compute grid of this size with this powerful network to do my thing on."
And so Nvidia's been our compute and network partners since they bought Mellanox for years now. And the thing that makes that work is generation over generation you're just getting better price performance from the systems. And we work super closely with them defining what the hardware requirements need to be in the coming generations of GPUs because we have a pretty clear sense of where models are going and what model architectures are evolving towards. So yeah, it's just been a super good partnership. We're deploying Hopper now at scale and a bunch of the features of Hopper like 8-bit floating point, arithmetic, and a bunch of other things or things that we've been planning for for a while.
Elad Gil: Yeah, I guess, one last question on this both supercomputer as well as platform side of things is I'm a little bit curious how you view the world shifting in terms of close source and open source models, and the mix that'll exist. Because obviously from an Azure perspective, lots of people are running open source models on top of Azure right now.
Kevin Scott: Yeah, it is an interesting thing that people are framing it as some kind of binary thing. I think you're going to have a lot of both. We still don't see any reason to believe that you're going to want to not build bigger models, but we just know in our own deployments, if you look at things like Bing Chat, or Microsoft 365 Copilot, or GitHub Copilot, you end up using a portfolio of models to do the work. And you use it for performance and cost optimization reasons, and you use it for just precision and quality reasons sometimes. And so there's always this melange of things that you're doing, and it's never either or. I'm actually really excited by what's going on with the open source community. I think my biggest question mark there is how you go deal with all of the REI and safety things. But if you look at the technical innovation inside of the open source community, it's really thrilling and we're doing some cool stuff right now.
I was just playing around yesterday with that 12 billion parameter Dolly 2.0 model from Databricks, which runs quite nicely on a single machine. And yeah, I'm still enough of a dork to love playing around with things that run on single machines. It's really, really impressive work.
Elad Gil: Yeah, yeah, yeah, it's super cool. How do you think about that from the context of enabling AI for your business customers outside of your core products? So is there a specific B2B AI stack that's coming? Are there specific tools coming? To your point, there's safety, there's analytics, there's fine tuning. There's so much stuff that you could potentially provide. I'm just sort of curious how you think about that.
Kevin Scott: Yeah, I don't want to turn this into some kind of weird marketing spiel, but we have this point of view that we started with this assumption that AI is going to be a platform, and the way that people are going to make most use of the platform is by building tools that assist people with jobs. So it's less about these fully autonomous scenarios, and more about assistive tech. And so the first thing that we built was GitHub Copilot, which is a coding tool. It's a thing where you can say in natural language what you would like a piece of code to do and it emits the code. And then you, as the developer, the same way that you would take a suggestion from a payer programmer, you scrutinize it, then code review it, and decide whether or not it makes sense for your application. And that was the first version of GitHub Copilot. It does a bunch of other things now.
And so the thing that we have observed is this Copilot pattern is actually pretty generic, and we built a bunch of Copilots since then. And the way that we built them, there's a Copilot stack that looks almost like one of these OSI networking diagrams, and it starts with a bunch of user interface patterns that you have. They're now an emerging plug-in ecosystem for how you extend the capabilities of a Copilot for things that you can't natively get out of the model. And then it is a whole stack of things, sort of an orchestration mechanism like LangChain is one of the popular open source orchestrators, but there are a bunch of open source orchestrators. We have one that we've developed called Semantic Kernel that we've also open sourced. There is this whole fascinating world right now that didn't exist nine months ago around prompt construction and prompt engineering.
And so there's an entire art form and a set of tools that people have access to design a meta prompt, which is the standing instructions to the model to get it to conform itself to the application context that it's in. You have these new things like new software development patterns like retrieval augmented generation or RAG, we were doing this before it had a name on it. So it's basically a way to take the prompt that's flowing from the application and to inject context into the prompt that will help the model better respond. And then there's a whole bunch of safety apparatus that you have. So that looks a lot like filtering on both the way down as the prompt flows through the stack all the way down to the model as well as it flows back up. So what things are you not going to let the application or the user send all the way down to the prompt because it's going to get a bad response back, or what things are you going to filter out at the last minute because it is a bad response that has gotten all the way through.
And sometimes you have multiple round trips through this cycle before you bubble the thing all the way back up to the user to get them the response that they need. And so we have a point of view about what this stack looks like, which Microsoft tools exist that will help people build these things, and what special things you have to go do in the context of an enterprise to answer the actual direct question, where safety and data privacy and understanding where the flows of data are and which plugins can be enabled and which can't. All of those things I think are getting built out right now. And the other thing too I'll say is, we'll build some of this stuff, and the community is going to build a tremendous amount of it because there's never been a platform or ecosystem where one company builds all of the useful things. That's just nonsense. It's just never happened.
And to me, it's the sort of super exciting thing to just see all of the energy that's happening right now. I just immediately before this call, I was doing a review with Microsoft Research, and it's just amazing to watch MSR, which is so many researchers there have pivoted what they're doing research on to these AI adjacent or AI on point things. And it feels a little bit like what MSR was like when I was an intern there in 2001 where you had all of these super bright people who had the tiniest little glimpse of what the future must look like that no one else had. Because it was the point where the PC was racing to ubiquity. And they were just all orienting their research around what that little glimpse was that maybe they had the earliest peak at. And it just is like, feels magical.
Elad Gil: That's massive realignment of the research community right now, sort of in real time. It's very exciting to watch.
Kevin Scott: And it's awe inspiring. I mean, it's just crazy. It's hard to keep up, like super hard. This has been the biggest surprise for me is I just didn't realize that GPT-4 and ChatGPT were going to catalyze as much of this as they have. We'd sort of been expecting a bunch of this stuff. ChatGPT was a-
PART 2 OF 4 ENDS [00:28:04]
Kevin Scott: ... kind of been expecting a bunch of this stuff. ChatGPT was a 10-month-old model with a little bit of RLHF on top of it, and by admission, not a beautiful user interface. It was just sort of a way to get something out there because you needed some practice with a handful of things before the big GPT-4 launch was coming, and no one really knew that it was going to blow up this way.
Elad Gill: And it's only five months old. That was only five months ago, which is shocking. I think everybody forgets how little time has passed.
Kevin Scott: Just shocking. But it is the open source community, and the big tech community I think at its best, is everybody is sort of realigning to what I think is unlike some of the other faddish things that have happened over the past handful of years. I don't think this is a fad, this is real.
Sarah Guo: I launched my new fund about six months ago with this AI focus and a few weeks later, ChatGPT comes out and I'd say even the people who are very prepared, hopefully somewhat prepared to go try to keep up or be part of that massive shift feel constantly upended, but it's the most fun time to be in technology in decades.
Kevin Scott: Look, it's also, I will say, a disconcerting time to be in technology because so many things are changing at once. It's changing at a pace that, you probably, even me, I think I might be in one of the better positions to feel like I'm kind of in control of what's going on, and I'm not in control at all of the pace. And so it must really be disconcerting to folks trying to keep up with everything that's going on. And in some cases it's forcing people to change their worldview about things, worldviews that they've held for a really long time. I think it's honestly harder for some machine learning people than it is for a brand new entrepreneur who's just looking for an interesting thing to go do because it is a very different way for a machine learning team to do its work, and it's been hard even for some of the people at Microsoft who have had plenty of time to think about the transition to get adjusted to this new way of doing things.
Sarah Guo: I want to ask you one more question that is advice for people making the adjustment in a certain sense. And then talk about your book, talk about the macro and such. Microsoft has a unbelievably wide portfolio of products and now you're on the other side of all the infrastructure questions, figuring out the organization of adoption of all these capabilities into that portfolio. I talk to friends who run large companies, started large companies all the time that are also figuring out how to do this, how do you organize that effort? What advice do you have for them?
Kevin Scott: I think you have to remember that some things have changed and some things haven't changed at all. And so one of the confusing things that I think there is for folks that many people get wrong is models aren't products and infrastructure isn't a product. And so you need to very quickly understand what it is this new type of infrastructure and this new platform is capable of, but that does not mean that you get to not do the hard work of understanding what a good product is that uses it.
One of the things I tell a lot of people is probably the place where the most interesting products are, are where you've made the phase change from impossible to hard. So something that literally you couldn't do at all before this technology exists has become hard now, because the things that have gone from impossible to easy are probably not interesting. And my frivolous example of this is when smartphones came on the market 15, 16, 17 years ago now, yeah, 2007, I guess, was iPhone launch, so 16 years ago almost. And then a year later you had the app store. So the first apps were things that had gone from impossible to easy and we barely remember them. There were all these fart apps. There was this app I had on my phone at one point that was called the Woo! Button, you pressed it and it did a woo like Ric Flair. Those are not businesses, they're just these explorations that people are doing.
The things that have made the smartphone platform are the hard things that went from impossible to hard. They also are the non-obvious things. They weren't even the things that the builders of the platform imagined. We don't even think the original applications on these platforms, the things that launched when the platform first launched, those are not the interesting things anymore. Your smartphone is way more than just a SMS app and a web browser and a mail client. The thing that makes it interesting is TikTok and Instagram and WhatsApp and DoorDash and these were all of these hard things that people had to go built now that they were possible.
And so I think that's thing number one to hold in your head, either as an entrepreneur or as a business that's trying to adopt this stuff. It's not like how I go sprinkle some LLM fairy dust on my existing products and do some stupid incremental thing. And I shouldn't even call it stupid, maybe the incremental things are fine, but the really interesting things are non-obvious and very not incremental. So that is the hard thing for us is you have an entire group of people who are smart and they can see all of the things that are possible, and so the challenge is to steer them towards the hard, meaningful, interesting, non-obvious things that are possible, not the things that are incremental that just are going to burn up a bunch of GPU cycles and a bunch of product IQ that will prevent you from doing the things that really matter.
Sarah Guo: If we sort of zoom out to non-technical audiences, you wrote a book in 2020, you wrote Reprogramming the American Dream. Can you describe who you want to read the book and what you hope they'll take away from it?
Kevin Scott: When I wrote the book, it was not for people like us. So the premise of the book is that I grew up in rural Central Virginia. My dad was a construction worker, his dad was a construction worker, his dad was a construction worker. My maternal grandfather ran an appliance repair business and had been a farmer earlier in his life. So the thing that was true for everyone who was in my life, neighbors, members of the community, is they're just smart, entrepreneurial, ingenious people using the best tools that they could lay their hands on to go do things that mattered to them that created opportunity for them and sort of solve problems for their communities.
And I believe that particularly this platform vision of AI where it's sort of getting cheaper and it's getting more accessible all the time, things, the stuff that we were chatting about few minutes ago about what I did at Google. I came in with a graduate degree, I was mathematically sophisticated and the first project that I did, which was a machine learning classifier thing in 2003, 2004, that was stacks of super technical research papers and elements of statistical machine learning, you read it cover to cover and then you go write code for six months. A high school student could do the whole project in four hours on a weekend now.
And what's happening, that aperture of who can use the tools is just getting bigger and bigger and bigger over time. And so the book was trying to get people to be inspired by this notion that don't be daunted and intimidated or scared by AI, go embrace it and try to plug it into the things that you're doing. And maybe we've got a shot at having more equitable distribution of who's benefiting from the platform as it emerges.
Sarah Guo: If you were going to add an update chapter for the last few years where so much has happened, what would you focus on?
Kevin Scott: Well, it's really interesting how much of it I think is still true. And I had this anxiety the whole time that I was writing the book that I was going to, by the time I had the manuscript in and it hit the presses that all of it was going to be out of date. The real problem I had is by the time it hit the presses, we had a global pandemic and it hit the presses the week that everything shut down, so you literally couldn't buy it. Amazon wasn't delivering anything other than essential packages and every bookstore in the country was closed. So it's a little bit surprising to me how many of the ideas that, we have a platform, the platform's getting more powerful, it's getting more accessible, actually the unit economics of it are getting better. What you can do for per token of inference is getting higher. So I know everybody's in this frenzy around GPUs which is this very expensive thing, but all of this optimization work is happening where you're able to squeeze more out of the compute that you have and the compute is getting cheaper.
So yeah, I mean the update that I would add is that, and it may be an update that I do, it probably won't be this book, but I'm sort of contemplating writing something right now. I do think that the public dialogue around AI right now is missing so many of the opportunities that we have to go deploy the technology for good. All of the articles that you read in the newspapers are around the responsible AI stuff, which is important, and the regulatory stuff, which is important, but we should have a few articles in there as well about Sal Khan's TED Talk, which is just amazing, unbelievably good. And just for folks who may not have seen it, which they should go see is, his problem is perfect for AI. So it's this Two Sigmas problem, this idea that students who have access to high quality individualized instruction performed substantially better than those who haven't, controlled for everything else.
Sarah Guo: Just for our listener's sake, The Two Sigma Problem was this study by a guy named Benjamin Bloom, which showed that your average tutored student performed above 98% of students in a control class, which is one teacher to 30 students, like a normal American classroom, with reduced variance, which is amazing.
Kevin Scott: Yeah. And if you believe that that's true, then you can also believe that every student, every learner in the world deserves to have access to that individualized high quality instruction at no cost, which seems like a reasonable thing. And then when you think about how you go realize that in the world... The only way that you can realistically do it is with something like AI.
And so there's so many problems that have that characteristic where we can all agree that it is a universal good to do this. And then if you think about how to do it, you must conclude that AI is part of the solution. That is the reason I get up every morning and deal with people yelling at me about, Give me my GPU's," for the fifth year in a row, it is because of things exactly like that. And it doesn't mean that when you talk about that and you're hopeful and optimistic about those things, or even hopeful and optimistic about all of the things that venture-backed companies are going to go do or the way the businesses are going to reinvent themselves, that you are also say, given the middle finger to the responsible AI concerns or the things that people care about on the regulatory front, you can care about both of those things at the same time.
But the thing that I can tell you is there is no historical precedent where you get all of these beneficial things by starting from pessimism first. Pessimism doesn't get you to optimistic outcomes.
Elad Gill: Yeah, it seems like, to your point, a lot of the dialogue is really lacking from global education, equity, global health equity, all these things that AI as a platform should be able to produce because it's cheaper, it's personalized, it can do things at the level of a human in many cases in terms of being a great teacher or a great physician's assistant, et cetera. And so it really feels like that message is lost. And I think a lot of people don't mention enough how we're almost hopefully going to enter this golden age if we let this technology actually bloom and be useful. I guess the...
PART 3 OF 4 ENDS [00:42:04]
Elad Gill: ... age if we let this technology actually bloom and be useful. I guess the question that I always have on my mind relative to all this stuff is given the capabilities that AI continues to accumulate, how do you think about 20 years from now in terms of the best roles for people and in particular I think about it in the context of my kids. I'm like, okay, normally two years ago I would've told my kids go study computer science. It's the language of the future. What do you think is the right advice to give people in terms of what to study and that will be the things that will be most durable relative to the change that's coming?
Kevin Scott: Yeah, I think... So 20 years is a tough time horizon.
Elad Gill: It's really tough.
Kevin Scott: And I think if any of us are honest with ourselves, if you rewind 20 years and you sort of imagine the predictions you would've made then, would you have gotten here? And nope, nobody would. But I think they're just some sort of obvious things. My daughter, for instance, has decided she wants to go be a surgeon, and I think surgeon is a pretty good job. We do not have a sort of robotics exponentials right now. We've got a cognitive exponential. And so I think all of the world is just full of these jobs where really affecting change on a physical system, doing something in the physical world, all of those things, we will need probably many, many more of them than we have right now. Particularly in medicine like nurses, surgeons, physical therapists, people who work in nursing homes.
We have a rapidly aging population, and so the burdens on the healthcare system are going to get much higher. And I do think that AI is going to have some pretty substantial productivity impacts, but maybe it's just enough productivity impact to make room for all of the other net new things that we will have to have there. And so I think we got this weird thing in the United States where we apportion less dignity and respect to jobs, the ones that my dad had than we should. And I lived in Germany for a little while. And Germany's a little bit different on this front. You can go be a machinist in Germany and that's a really great career and something that your parents are celebrate.
So I think they're all of these careers, electricians and machinists and solar installation technicians. And I mean, just so many things that we're going to need, especially because we're going to have to rebuild our entire power generation and distribution system in our children's lifetimes. So all of those jobs I think are super important. And then I would argue even that all of the creative stuff like that we do, there's going to be probably more need for that in the future than less. Even though the tools that we're using to do the creative work, whether it's coding or making podcasts or whatnot, are going to help us be better at it. And the reason that I say that is humans are just extraordinarily good at wanting to put humans at the center of their stories.
Right now, we could be making Netflix shows, not Queen's Gambit, but machine's gambit about a fleet of computers playing chess among themselves because they're all better than the very best human. Nobody wants to watch that. The technology's probably good enough right now where you could have superhuman Formula One closed track drivers and Formula One cars that could do things that humans can't do. Nobody wants to watch that. And you even go back before computers. Forklifts are stronger than people. You could go have a strong man or a strong person competition that was about which forklift could lift the most weight. Nobody cares about that. We care about humans. What are we saying? What do we care about? What are we trying to express to everyone else? And nothing about that's going to change nothing.
Elad Gill: I think that's why people watch the Real Housewives of Dubai.
Kevin Scott: Yeah. Again, I don't want to paint too rosy a picture. Every time you have a major technology platform or paradigm shift, there's disruption. But what we know from every one of these disruptions is you have actually a surprising degree of need for human occupation. All of the Industrial Revolution predictions about four hour work weeks, and we're all going to live lives of leisure is bull crap. Just hasn't happened. And I think some people may say it hasn't happened because the system doesn't want it to happen. But I think a lot of it is because we actually like doing things.
Elad Gill: And there's a lot to do. I guess on that note, what are some of the areas you're most excited about going forward in terms of the coming year of AI or big research areas or big product areas or things that you're very optimistic about?
Kevin Scott: I think two things, just I think this will be maybe the great first year of foundation model deployments where you're just going to see lots and lots of companies launched, lots of people trying a bunch of ideas. You're going to see all of the big tech companies will have substantial things that they're going to be building. I got predictions about what other folks will do, but it will touch all of Microsoft's product portfolio. The way that you will interact with our software will be substantially different by the end of this calendar year than it was coming in. And I think that'll be true for everyone. I think it changes some of the nature of the competition that you've got between big tech companies, and I think it creates new opportunities for small tech companies to come and drive wedges and to get footholds and do interesting things.
One of the things that Sam Altman and I have talked about a lot is I suspect that this year, the next trillion-dollar company gets founded. It won't be obvious which it is, but we're overdue, long overdue. And then I think what you're going to see technically this year is I do think that you will have things like the Red Pajama Project is like this... And there're going to be a bunch of others like it will make really good progress on building more capable open source models, and hopefully the community will help build some of the safety solutions that you will need to accompany those things when you deploy them. But technically, I think you're just going to see amazing progress there. And then the frontier will keep expanding out. Open AI doesn't have GPTV in wide distribution, but it'll get to wide distribution at some point in the not too distant future.
And so you'll have these very powerful multimodal models the same way that having GPT4 admitted all of this exciting energy around new things that you could do with it. Having a model that can take visual inputs and reason over them will also admit a whole bunch of new things that are going to be very exciting. So I don't know. I just think that the theme of this year is going to be progress and activity almost too much to track. I'm going to need a co-pilot just to pay attention to all of this stuff and make sure I'm not missing important things because I feel like I'm at the... And you all as investors and as people who are watching this closely must feel the same thing. It's like, how do I make sure I don't miss the next important thing? How do I see it as soon as humanly possible?
Sarah Guo: Actually, just to make completely sure, if you are starting the next trillion-dollar company in our listener base this year, please call me in [inaudible 00:50:41] and Kevin, too. Wrapping up now, is there anything else you would want to touch on, Kevin?
Kevin Scott: I think the dialogue that we're having right now around regulation is actually really quite important. So as we're recording this, Sam Altman was testifying in front of the Senate Judiciary Committee on Tuesday of this week. I think more of those conversations are a good thing. I think as fast as things are moving, you really will need the technology community to come together and to agree on some sensible things that we can do before the regulation even is in place. And I think that's all important and not a thing. The thing that none of us should be doing at this point is looking at the prospect of regulation and saying, oh my God, this is a pain. I don't want to deal with this. The fact that there is a desire for it is a very good signal that the things that we're working on actually matter, because nobody's trying to regulate frivolous things.
And the purpose of regulation is to make sure that you can build a solid, trusted foundation for things that maybe become ubiquitous in society. If you think of this electricity, for instance, you want to strike the right balance between allowing the technology to develop and make progress and flourish, but you also need to make sure that your electric power generators are built safely and you don't allow people to wander in and stick their finger on the electrode and disintegrate themselves. And you want to make sure that the distribution of electricity is coordinated and that when it comes into your house, it doesn't burn your house down. And when you plug your appliances into the wall that they function as expected. And so I think that is a similar way. There's not going to be one size fits all. I think the most of the stuff that people ought to be thinking about is deployments.
Making sure that as you deploy the technology, getting the requirements and the expectations right there is the most important thing. And then these big engines that we're building that are the largest of the foundation models, making sure that you have a set of safeguards around those. But also the way that we are building these things, they don't get distributed to the world by themselves. There's a whole layer of things on top of them to render them safe. And then a whole set of things per application, per deployment that we do to make the deployment safe. And so I think everybody, all the startups, everyone in the open source community, everybody ought to be thinking about these things. How am I doing my part to make sure that we are creating as much space as possible for these optimistic uses, and we are deterring as many of the harmful ones as possible?
Elad Gill: Yeah, I've been impressed by the degree to which the community has self acted from very early days in terms of AI safety and a purchase to that. And so I know OpenAI's done stuff really early. Philanthropic has. Google has. Microsoft has. I feel like a lot of the main players have actually been remarkably thoughtful about this area and keen to make sure that it's done properly.
Kevin Scott: Yeah, I mean, the thing that I will say is we fiercely compete with a whole bunch of these folks, but one of the things that I don't do is look at any of those companies that you just named and worry that they're going to do something that I take myself out of my role as CTO of Microsoft and just think about Kevin's Citizen of the World. Kevin's Citizen of the World is not worried about what my competitors are going to do to do something unsafe. I'm just not.
Sarah Guo: Thanks so much for being with us, Kevin. We really appreciate it.
Kevin Scott: Yeah, thanks for inviting me. This was awesome.
Elad Gill:Yeah, thanks so much for the time. That was great.