Behind The Tech – Ask Me Anything with Microsoft CTO, Kevin Scott
CHRISTINA WARREN: Welcome to "Behind the Tech". I'm your co-host, Christina Warren, Senior Developer Advocate at GitHub. KEVIN SCOTT: And I'm Kevin Scott. CHRISTINA WARREN: It is time now for our AMA episode. And so for the past couple of months listeners have been sending in some really fantastic questions.
And we cannot answer every single one that we got, but we are so appreciative of all of you who sent in your questions, and so this is going to be a super interesting conversation. Here's our question from Ravinder, "How has your pace of learning changed in the era of AI? What's been the coolest thing you've done with AI personally?" KEVIN SCOTT: Yes, I've definitely been using AI a ton for the projects that I'm doing outside of work even, so like the bunch of things that it gets used for, for at work that are hugely useful. But the outside of work ones I think are kind of fun.
So like maybe the coolest thing that I've done is I have gotten really into making Japanese tea bowls in my ceramics studio this past year, and I have been researching how to replicate some of the results in traditional, classic Japanese Raku tea bowl making, which has involved me making my own kiln, devising my own glaze recipe, and even devising a way to take a clay body that you make the bowls out of and making it tougher so that it can handle all the thermal cycling and this crazy firing process. And I will tell you that Copilot was amazingly useful in all of that, like particularly with the kiln design and with helping get some ideas and make progress on the glaze chemistry for this glaze. CHRISTINA WARREN: That's so interesting. So like what do you do with Copilot that you just -- or do you just have a conversation and just ask questions kind of back and forth about maybe how you want to design stuff? KEVIN SCOTT: Yes, I mean, I -- basically for the glaze design I told -- or I asked Copilot, I was like, "I've got a set of tea bowls that I am firing in the classic Raku style at 1100 degrees Celsius where I'm going to take the glazed vessel, put it directly in at temperature kiln, like leave it for three minutes until the glaze goes cherry red and then pull it out to air quench. And I gave it a few hints about what I had been thinking about -- like what I know of the Japanese -- like this is the interesting bit, the traditional Japanese Raku glazes use lead in them to get the elements of the glaze to melt at a lower temperature.
CHRISTINA WARREN: Ah. Okay. KEVIN SCOTT: And obviously I don't want to be using lead in my tea bowls, even though there are like safer variants of lead that you can use that are safe in ceramics; but I didn't want to, and so you need to use something else like boron. And so figuring out how much boron you use and in what form the boron comes in in the glaze is like a little bit tricky. And like it was super helpful. And it was -- it felt like a real conversation that I was having with someone who knew a little bit something different about glaze chemistry than I know.
CHRISTINA WARREN: But that's -- genuinely, this is so fascinating. And also, thank you for sharing all of your many interests with us because you're such an interesting person. And I never would have thought like that from all -- I knew how much of the maker stuff you're into, but building your own kiln and making Japanese tea bowls, and using the AI to get more information about this, I love it.
That's -- KEVIN SCOTT: Yes. CHRISTINA WARREN: -- a great use case of AI. I love that. Great stuff.
Thank you again for that question. All right. KEVIN SCOTT: Yes. CHRISTINA WARREN: This question is now from Rafael, and he asks, "Do you believe that in the future AI will completely reshape the way that we produce software?" And he goes on, "Could we eventually get rid of the development tools that we use today and rethink the entire process from scratch, creating a completely new approach to software development?" KEVIN SCOTT: Seems likely. CHRISTINA WARREN: Yes.
KEVIN SCOTT: I mean, and I'm an old enough fart where -- I mean it sounds disturbing to say, but like I've been programming for 40 years. So I'm 52, I started when I was 12. And in 40 years, like particularly the past 40 years, like it -- software development now even without AI doesn't really resemble much at all what software development looked like in the 1980s. And so I think it's a safe bet that software development is going to reform itself over the next handful of years, and I'm, I think, just super clear that AI is going to change the way that we write software.
I think -- yes there are just sort of all of the obvious ways that it's going to change things. Coding is a complicated activity, and like it always has been like this thing where you've got an idea in your head that needs to be sharpened and then you need to get the sharpened idea out into a form that the computer can go execute. The thing that's really changed -- and I've said this before in public I think, is the way that we've been building software hasn't really changed since Ada Lovelace; like this whole -- CHRISTINA WARREN: Right. KEVIN SCOTT: -- process or algorithmic thinking and understanding the complexity of a machine like all the way down to like its atomic details and then using that understanding of the machine to like transform this idea that you framed algorithmically into a program that the computer -- right, like we've been doing that for almost two centuries now, and there really hasn't been much of an alternative. Like our tools have become increasingly more powerful, but it's like it's basically that you want a computing device to do something for you, you either like figure out how to do that process yourself, or you have to hope that someone who understands how to do that has written a program that you can run yourself.
And I think the big thing that's changed with AI is now you have a thing where you can describe a thing that you want accomplished, not necessarily even in algorithmic terms, and then the AI can do some or all of that mapping to get the computer to actually do the thing for you. And like that really, really dramatically changes how we think about software development and who's the developer. It changes like what it means that we're building. So for instance, I was just having this conversation with a bunch of engineers like, "I don't know that you need apps in this world." So like an application is a byproduct of this early thing that I just described, that someone has to understand a set of problems that a group of people want to accomplish and then they just sort of added a bunch of code together into this thing called an "application" that does those things in a general enough way that those people can get some value out of it and be able to use it.
And I don't know that you are going to need that too much further in the future. Like you'll still need like the capabilities that are in the applications, but the user interface like telling someone they've got to go learn all of the complexity of some software because they've got to navigate some weird user interface -- CHRISTINA WARREN: Right. KEVIN SCOTT: -- information architecture to get a thing done, versus just say what they want done, like that's changing clearly. And like that has an implication -- CHRISTINA WARREN: Yes.
KEVIN SCOTT: -- for the software development as well. CHRISTINA WARREN: Yes, yes, it does. I mean, that's what I kind of think about, right, because obviously I think you're right, like it could change completely how we define a developer, which is something that we've been trying to do in various ways for a long time, but now we finally feel like we're maybe on the cusp of really broadening that concept; like it really feels like that could be your reality.
But it does make me think about on other levels, "Okay, so how do you design programming languages, or do you, or how does that change, right, like -- KEVIN SCOTT: Yes. CHRISTINA WARREN: -- what matters then about the underlying code beyond that if we are able to just create things based on our natural language and based on what we want and make updates iteratively with multiple people at once, like how does that change how we design those underlying systems?" I think that's really interesting to think about, too. KEVIN SCOTT: Yes. Yes, hundred percent; and super useful stuff to think about. And like the trick is -- and this has been true about software development forever, is like you want things to compose.
So AI is still a pretty far ways away from doing this grand vision that I just articulated. And so like what we really need to be thinking about between now and like whenever that happens, if it actually happens the way that I imagine, is like how do you take tools that are on some spectrum of classical software development tools to this new AI future and make sure that all of the things compose together in reasonable ways so that developers can then take all of the stuff that's in their toolkit and like get the thing built that they are hoping to be able to build. CHRISTINA WARREN: Yes. Yes, totally. I mean, lots of things to think about, and I totally agree with you on that.
All right, we've got this question from Veronica, shifting just a little bit, still around AI. And she wants to know, "How do you suggest we regulate AI? Should this be done at the federal or state level, and how can we ensure that AI is safe and secure, with both public and private standpoints?" Great question. KEVIN SCOTT: Yes, I think it's a super great question. Again, I've said this as well, and like I think I talked about it even some in my book.
Like of course any technology as powerful as AI needs to be regulated, and it would be just an odd thing in the course of human history if you had something this powerful and it wasn't regulated. The thing that you want to do, though, with regulation is I think consistency is helpful, like that's where sort of federal regulation that is consistent across all the states and even sort of international standards would be super, super useful; because regulation -- like good regulation's intent is to like get beneficial technologies deployed to those who will benefit from it as quickly and safely as humanly possible; and so you don't want unnecessary complexity in the regulation itself, because like that prevents the whole beneficial technologies getting to whom it benefits. But yes, I mean I think in general we will need our regulators to be pretty agile in making regulation that can encourage the most beneficial things for the broadest number of people to get to the market as quickly as possible while at the same time being careful about what the downside risks are to a bunch of things. And in a bunch of places like the biggest downside risk, honestly, is failure to deploy quickly enough. CHRISTINA WARREN: Mm-hmm. KEVIN SCOTT: Like there are, for instance, like a whole bunch of medical things right now where the models are strongly superhuman.
And I've had some experience with my own mother in the past year with the healthcare system where if she had had access to the most advanced AI tools a whole lot of suffering could have been reduced. It's lots and lots and lots and lots of people are in similar situations where it's not some theoretical future where stuff could be beneficial, it's now that it could be beneficial and -- CHRISTINA WARREN: Right. How do you think we go about I guess educating or ensuring that our legislators are aware of what the potential I guess what opportunities and risks are in this area; right, because this is something I think about a lot. I agree with you, regulation is super important, and it needs to be consistent, but I do sometimes wonder -- I mean, it's hard enough for us as technologists to keep up with all of these things. How can we do a good job of making sure that the legislators are informed? KEVIN SCOTT: Yes, I will say the thing that I'm most encouraged by on this front with AI is more so than any previous technology that I'm aware of; you have practitioners in the field spending a whole bunch of time talking with people in the academy and people in government trying to make sure that they have the information that they need in order to make good decisions.
And I see people doing it in very respectful ways. Now, obviously everybody who's coming at it like whether you're in the government, or you're in the academy, or you're in the industry, like you're obviously biased in some way. And so -- CHRISTINA WARREN: Sure. KEVIN SCOTT: -- we all need to be as clear as we possibly can about our biases and sort of lay them on the table. But like just because you're biased like doesn't mean that you can't get information out there and then have someone adjust for the biases, look for what the through line is and everything and then make good policy decisions.
Like that's a way better way to be than to like not be transparent about what's going on -- CHRISTINA WARREN: Yes. KEVIN SCOTT: -- or decide that you're not going to talk to somebody because it's not your job. Like I think right now in tech anybody who's working on AI like part of your job is to when required patiently explain what it is you're doing, why you're doing it, and how it works. CHRISTINA WARREN: Great stuff. All right, so this is a question from Muhigiri, and this is really good, "How can large language models be scaled effectively across regions with limited technological infrastructure?" So think about places like African nations, like what are some of the biggest hurdles for AI-powered educational solutions to move beyond prototyping and into full-scale deployments in underserved regions, and how can these challenges be overcome? KEVIN SCOTT: Well, I think the news there is probably pretty good. So if what you want to do is to build an AI application, it has never been easier than it is right now to go build one.
You have more choices about very powerful models to access. You have models that are available behind APIs that are hosted where you sign up for a developer key and just start making requests. You have like a huge catalog of open-source models that are on a spectrum from general purpose to like very specific task design things. And so like you just have a lot of choice where you don't have to start by saying, "I've got to train a model from scratch." CHRISTINA WARREN: Right; right. KEVIN SCOTT: And so I think that is a huge advantage.
Like it's definitely not the way things were 20 years ago when I wrote my first machine learning programs, and it isn't even how things were three or four years ago, where like -- CHRISTINA WARREN: Right. Right; I was going to say it's a lot different even than then, right, it's much easier for people to build really good things now versus three or four years ago, to your point. KEVIN SCOTT: Yes. I mean my boss, Satya Nadella, tells stories about his visits to India recently where he has seen the just rapid diffusion of AI applications at a pace that he's never seen before.
Like the thing that he says which I think is really good is there are parts of rural India where the industrial revolution still hasn't shown up after 250 years, where they already are seeing the diffusion of AI, where a farmer through their mobile device can access a powerful AI system that will help them understand how they are entitled to government programs and then go sign them up for them so that they get these benefits that their government intended them to have. And like that's just kind of a shocking rate of diffusion. But look, it's also not all good news.
Like I think while the expertise required to build an AI application is democratizing super fast and you've got like high levels of accessibility to the APIs and basic infrastructure required to go build them, you still have to be connected, you still have -- CHRISTINA WARREN: Right. KEVIN SCOTT: -- to like have some baseline level of technology fluency in order to be able to use the systems. And like the reality is there are large parts of the world that are not yet sufficiently connected and where like the technology fluency isn't as good as it should be. And so I think there's a bunch of at this point deeply unsexy work that we still need to prioritize and make sure that we're focusing on things like just rural broadband.
CHRISTINA WARREN: Mm-hmm. KEVIN SCOTT: Like I've definitely told this story before, but my mom and brother have good internet in this rural town that they live in, in Central Virginia because they are lucky enough to live within a hundred yards of the local telco exchange. My uncle, who lives just a few miles away from them is still on some kind of like crazy 300K DSL connection, and his internet is barely usable. And so he has to come to my mom's house to do things on the internet. It's nuts.
And so like that's the sort of thing that I think we really have to pay attention to, because as the things that you can do and the capabilities you can access with that connectivity become more powerful, like absence of connectivity like -- CHRISTINA WARREN: Yes. KEVIN SCOTT: -- becomes a bigger and bigger disadvantage. CHRISTINA WARREN: No, I mean, I think you're exactly right. And this is a conversation I feel like -- and we've definitely talked about this on this podcast, but I feel like we collectively as an industry and society have been talking about this for at least 20 years, and it's only becoming more and more important, right, to start to really invest in overcoming these infrastructure challenges just because connectivity is only going to be more important. Right, I think that's a great distinction, that it's easier than ever to build applications and things with these tools, but actually getting it to people and making it so that they can interact with them is the -- maybe the less fun part, but arguably even more important because -- KEVIN SCOTT: Yes. CHRISTINA WARREN: -- without that we -- all of this is moot; yes.
KEVIN SCOTT: Yes. CHRISTINA WARREN: All right, question from Peter. He asks, "I am curious about how Microsoft approaches running technical tests against its own infrastructure, LinkedIn, Xbox, Office 365, and others. Given the scale and complexity of these systems, how many lessons have you learned over the years while managing that infrastructure?" And he goes on to ask, "And for those of us in DevOps, what is the most surprising lesson that you've encountered that might catch us off guard?" KEVIN SCOTT: Oh, God, that's a super good question, like very complicated -- CHRISTINA WARREN: Yes. KEVIN SCOTT: -- so I don't know whether I'm going to be able to answer the whole thing.
CHRISTINA WARREN: I was going to say -- KEVIN SCOTT: Like I think -- CHRISTINA WARREN: -- if you need to take this in parts, do that, do that; that's okay. KEVIN SCOTT: Yes. Look, so I had a boss who was like maybe the best DevOps leader I've ever worked for or with in my career, and like he had a bunch of like very simple things that he would say about philosophically how you should approach DevOps. Yes, like one of the things he says is -- or said is, "You can't fix something or improve it if you're not measuring it." So a lot of the answer to the question it just boils down to like are your metrics good, are you measuring everything that's happening in your system, do you have good monitoring built on top of the metrics, like do you have good visibility into the internal state of all of the systems; like that's one thing that's super important.
Another thing is like complexity needs to have a reason. And so a lot of times complexity just sort of emerges because like the most convenient thing to do to systems architecturally is often just to pin new stuff onto old rather than to do the harder work of, "Okay, like we've got some evolved requirements here, things are different from when we originally designed this system; like now we need to like just push pause and go refactor the whole system and make sure that it's designed in the simplest possible way to meet the new set of requirements that we now understand." And so like one of the things that I've always tried to do in the organizations that I've led is to make sure that you are reserving some amount of your engineering capacity to go deal with tech debt, that you've got teams who are building shared infrastructure whose job it is not just to provide a set of services to everyone, but to like be building things in a really architecturally simple way, and to like make sure that things are robust, maintainable, scaleable, secure, fault-tolerant, like all of the things that you want out of your systems.
And like you've just got to rebuild stuff every now and again. Like it's painful as it may sound, when you've got product managers screaming at you that like, "you need to go ship this new feature," or you're eyeballing short-term revenue or something like that to like go tell all of your stakeholders, "Hey, we've got to push pause on this for a little while while we like rearchitect this thing," you just have to do it because complexity really is the -- it's the killer. There's a bunch of stuff that we're doing with AI right now, though, to deal with some of the complexity, so like it can -- when you have complexity in systems, it's irreducible, like you just can't figure out how to design away from it, like AI can help manage some of the complexity. And it's like not in a way where you're letting this AI be an abstraction layer that sits between you and your understanding of your system, but to like help you just -- CHRISTINA WARREN: Right. KEVIN SCOTT: -- very quickly like triage things or figure out like how to root cause operational issues or whatnot. It can be super helpful with stuff like that.
Yes, I mean, like I could go on all day about this particular bag of issues, but yes, I mean, like you've just got to test, test, test. And like -- here's -- I think maybe this is -- I have gone into situations before where people have built systems or built functionality that are designed to do a thing in rare circumstances like sort of data center level fault tolerance, for instance, so like what happens if this whole data center goes down, if it loses power, or if like there's a fiber cut or something where the team tests the functionality once, and then assumes that it's going to be available forever and ever just because -- CHRISTINA WARREN: Right. KEVIN SCOTT: -- it worked one time.
CHRISTINA WARREN: Right, right. KEVIN SCOTT: (Laughs) And so yes, you've just got to -- you've got to test for infrequently occurring things and make sure that when the infrequently occurring thing happens that you are ready to go, which basically means you need to simulate the infrequent thing more frequently than it will naturally happen. And that's like a counterintuitive thing, I think, for some folks. CHRISTINA WARREN: Yes, no, that is; but I like that.
I think that probably answers the question really well because that does seem counterintuitive, but it makes sense, right, like you need to make sure that when this actually occurs that it's going to work. But to do that, got to have -- it's kind of like fire drills right, like, you do them -- KEVIN SCOTT: Yes. CHRISTINA WARREN: -- hopefully much more frequently than they actually occur, just in case if you need to be ready. KEVIN SCOTT: Yes; no, I was just going to say, like at LinkedIn we used to at random points every week just take a whole data center offline to make sure that all the fault tolerant systems would work. CHRISTINA WARREN: (Laughs) Okay, that's awesome.
That's wild. And was that a process that started before you joined or was it something that you asked them to do? I'm just curious. KEVIN SCOTT: That was a thing I asked them to do. CHRISTINA WARREN: Amazing. Amazing. And was it for that reason, just because you wanted to ensure that -- KEVIN SCOTT: Yes, it was -- CHRISTINA WARREN: -- there was resiliency? KEVIN SCOTT: Correct.
It was because resiliency is a super hard thing to achieve, so it is not a service that you can just sign up for and get resilience. CHRISTINA WARREN: Right. KEVIN SCOTT: And basically means that every single thing that's running in the data center has to be resilient.
It has to be prepared to deal for things to fail in the worst possible way; which means like obvious things for things like databases, and networks, and storage systems, and whatnot, and like there's a bunch of super classic computer science and engineering stuff that you can go do to make those things fault tolerant. But you also have to make your applications fault tolerant. CHRISTINA WARREN: Yes. KEVIN SCOTT: Like what happens if like an application server that's rendering the user experience to a user, like what happens if it loses like all network connectivity; like what happens then? Is there some routing layer somewhere like maybe in the end user application that the user is using that will notice that its connection back to its application server is no longer responsive and it routes it sideways to another server somewhere in the service catalogue in another data center? It's like you just have to think through all of this stuff, like how is every single piece of this system -- and you have to have every single service owner accountable for having done that work. And like a real good way to make sure that they've done the work is without telling them, you just kill the whole system -- CHRISTINA WARREN: You just kill it.
KEVIN SCOTT: -- and like you'll know really quickly whether -- CHRISTINA WARREN: Yes. KEVIN SCOTT: -- their application's robust or not. (Laughs) CHRISTINA WARREN: I love it. I love it. And I'm glad you implemented that. I mean, I think it's a testament to LinkedIn that it is one -- I've covered many of these services and worked at companies that need to be online a lot that have not always had great enough time.
LinkedIn is one of the ones that has -- at least in my experience, a very, very good uptime and those sorts of things. And I think that's probably a testament to the drills that you ran. KEVIN SCOTT: Now, but not always.
(Laughs) CHRISTINA WARREN: Yes. Well, but that's how you get there, right, I guess is by -- KEVIN SCOTT: Yes. CHRISTINA WARREN: -- having to just -- at the drop of a hat, it could be gone. How are you going to recover? KEVIN SCOTT: Yes. CHRISTINA WARREN: I love that.
I love that. All right, this question is from Samantha, and she asks, "I've noticed you've had a few recent guests that aren't typical technologists like Ben Laude and Refik Anadol. Could you share more about your thinking and perspective on how more creative leaders are working in the era of tech and AI? KEVIN SCOTT: Yes, like a part of it is like just to be perfectly honest, like these are people that I want to talk to -- CHRISTINA WARREN: Yes. KEVIN SCOTT: -- and I think the conversations are interesting and I want to share them. But I think there is this thing that we had been talking about, which in the era of AI this distinction between like who's a technologist and who isn't is like blurring at a really profound way. And so I think it's good to be talking to a broader variety of people because you have -- like Refik, for instance, is a trained artist, but he's using technology in incredibly sophisticated ways to realize this artistic vision that he has.
And I think there's just going to be more, and more, and more of that over time because this previously daunting and inaccessible technology is becoming less daunting and more accessible and which means that more people are going to be using it to do a broader swath of things. And so yes, like is Refik an artist or a technologist, like maybe it doesn't matter. CHRISTINA WARREN: Right. KEVIN SCOTT: He's just doing amazing stuff. And like this conversation I had with Ben Laude is -- like I think all the time about what the nature of art is and like what are the things -- what's the difference between art and instrument, and like what's the boundary between performer and instrument? CHRISTINA WARREN: Mm-hmm. KEVIN SCOTT: And so I think it's interesting to have artists come in and talk about how they are thinking about those relationships that they had in their art and in their craft for a very long while, and then how that thinking is changing in an era of AI.
So I don't know, I just feel like there are super important conversations to have right now. CHRISTINA WARREN: No, I think you're right. And I think that kind of breaking down maybe kind of this demarcation in places -- like I don't know if it matters with yes, right, that could be the answer to both questions. And because these lines when technology truly becomes accessible and kind of is something that we all sort of kind of imbibe, it's not -- it becomes just a part of us, right, like it -- and I think that the oftentimes artificial barriers that we put into place disappear and it's just like you're a creator, you're a person -- KEVIN SCOTT: Yes.
CHRISTINA WARREN: -- regardless of how you get there and what you do; but it doesn't have to be a lot, "I have to be in this box or this box." It's like, "No, I'm just good at reading." KEVIN SCOTT: Yes. The thing that I will also say is like I have super strong opinions about some things. Like for instance, like I'm not interested in AI at all, absent a human wielding the AI -- CHRISTINA WARREN: Right.
KEVIN SCOTT: -- to do something interesting. Now -- CHRISTINA WARREN: Right. KEVIN SCOTT: -- I'm not claiming that everybody needs to be my way, but like it's just interesting to me that like this isn't a -- like a point of view that I came to through some like huge process of deliberation. It's just like I am not interested in the idea of some autonomous AI like spitting out art, or music, or whatnot -- CHRISTINA WARREN: Right. KEVIN SCOTT: -- absent the hand of a human creator; because like I've sort of discovered like part of my connection to the experience of -- or like of experiencing art in the first place is like I like to know like, "Oh, this is the human, and this is how they made it, and this is like imagining what they must have been thinking, and like -- CHRISTINA WARREN: Yes.
KEVIN SCOTT: -- are we alike, are we different?" And like that's part -- CHRISTINA WARREN: You like the story. KEVIN SCOTT: Yes, I like the story. And the story -- CHRISTINA WARREN: Yes; no, I think -- KEVIN SCOTT: -- of like, "The robot made this; who cares?" (Laughs) CHRISTINA WARREN: Right.
No; and I think that's a great point, right, and that's a really interesting perspective, because obviously -- I mean, I think there's an argument to be made that there is something artistic that could be made if it were completely autonomously generated. KEVIN SCOTT: Yes. CHRISTINA WARREN: And that's an interesting thing to have. But I tend -- KEVIN SCOTT: Yes. CHRISTINA WARREN: -- to agree with you, like ther stuff that I'm interested in consuming the most outside of kind of like an abstract level is definitely the stuff that has been guided by a human.
But if the technology, if the AI can make things more unique, or effective, or just add a different nuance to something, that can lead to a great outcome, so -- KEVIN SCOTT: Yes. CHRISTINA WARREN: -- or interesting anyway, so yes. But I like that perspective. KEVIN SCOTT: Yes. It's an interesting debate because I don't know whether I'm right, or you're right, or -- I do actually have this argument with people who like will say like, "Hey, you're crazy," like, "the -- you could have something that's interesting, and artistic, and merit worthy that doesn't have -- " and it's like, "Okay, great." And like -- CHRISTINA WARREN: Yes.
KEVIN SCOTT: -- the argument is interesting, right, like it sort of tells us -- CHRISTINA WARREN: It is; it is. KEVIN SCOTT: -- it tells us something about what is the nature of these things. CHRISTINA WARREN: No, I think it does, right, and like -- yes, because I can see both perspectives. I tend to, I think, align more with you, but I can understand like the philosophical argument about it. But I think that for a lot of us still what ultimately binds us to things is not just the output itself, but the -- everything that comes before it, which is the story and is the thinking about what went into it, and is frankly in some cases the imperfections, right? And that is something that -- not to say that it couldn't be there, because who knows where AIs might be in decades, but that's not -- that doesn't seem to be the direction that a lot of those things are now. And so instead, though, I think it's interesting to think about like how these tools can be used not to just clean up imperfections but to maybe continue to let those things be there but maybe show off other ideas, I don't know.
All right -- KEVIN SCOTT: Right. CHRISTINA WARREN: -- this question is from Kathleen and she says, "I've been hearing a lot about agents being the next AI frontier. What can you tell us about what that will look like and when can we expect to use AI in that capacity?" So great question, we all want to know when are the AI agents going to be able to run our lives, Kevin? KEVIN SCOTT: (Laughs) I don't know for sure. But like I -- so I think it's like important to be more specific about what it is we think agents are. So in a way, like copilots are agents, but they are sort of agents that can help you with, relatively speaking, small task.
There might be a lot of them that you're doing and they may be very important, but right now like the things that we can delegate to AI are relatively small, like small software development tasks, like small productivity tasks. And like what -- eventually like if you are excited about this notion of agents, what you want to be able to do is just sort of think about an agent as like a real fully capable peer, or collaborator, or coworker, and like you want it to be able to collaborate with you in like very broad and very capable ways, or you want to be able to like delegate like big things, so like for not just five-minute tasks, but five-day tasks like, "Go completely autonomously build this whole application for me and come back with the -- with a PR you want me to review and something that I can test," which you might do to one of your fellow software developers, right? CHRISTINA WARREN: Right. KEVIN SCOTT: And so, look, I think we're definitely moving in the right trajectory to like have these agents, which in our parlance we call "copilots", become more and more powerful and capable over time. So I think we're feeling really good about reasoning capability. We are beginning to make progress on actions in tool use, like we've seen a little bit of that in the past year, and I think you're going to see a bunch of it in the coming year. We are seeing like really interesting things happening, I think, and like have a lot of things that we can expect to see in the next year on memory.
Like a lot of what happens now with these agents is they're very transactional, so like they -- CHRISTINA WARREN: Right. KEVIN SCOTT: -- have enough information to do a very specific task in a very specific context. But in order to have them be more generally powerful, like they have to like really have complete memories and -- CHRISTINA WARREN: Mm-hmm. KEVIN SCOTT: -- persist over time. And then like we've got a whole bunch of plumbing work to go do. Like you -- in order for the agents to be able to do things like just even beyond basic tool use where they can take action on your behalf or where they can go use a tool to assist them in accomplishing the task that you've set them off to go do, like you really do have to think about like what do entitlements looks like in this universe, like how do you make sure that the agent has access to what it needs to have access to in order to complete the task it's been asked to do, and like how do we the humans reason over those entitlements and get things both available and permission correctly? Yes, so but look, I think I'm seeing lots and lots of progress, and like it's hard to predict like the date when agents with capability level X is going to be there.
But I think it's safe to assert that we will see increasingly powerful agents in a variety of different forms emerge over the next year. CHRISTINA WARREN: Sounds good. Sounds good. I think that's probably a good hedge, and I look -- I do look forward to the day that like the robot overlords do truly control my life. But until then -- (laughter) yes.
I'm kidding, I'm kidding. I'm kidding. But I'm glad to know that progress is being made. Okay, that does it for our AMA episode.
Thank you again so much to everyone who sent in these excellent questions. Really, really good stuff. Thank you, Kevin, for your answers; really, really interesting. Please make sure to follow "Behind the Tech" on YouTube or wherever you listen to podcasts. And if you have anything that you would like to share with us, you can email us anytime at behindthetech@microsoft.com.
Thank you so much for listening. KEVIN SCOTT: See you next time.
2025-02-21 06:48
