AI Impact | Christopher Young and Martin Migoya | Globant Converge

Absolutely. First of all, thank you, Martin, and to the entire Globon team for having me. I spent a lot of my career pre-Microsoft. I've been at Microsoft about four years now, and I lead our business development team, strategy, and ventures. We do a lot of investing work with partners. You may have heard of OpenAI as an example, all the way down to many, many partners. McAfee had a long history. The company had been acquired by Intel, and they had it as part of the company for a number of years. When we spun it out, we basically had to rebuild the company. Everything inside the company had to be rebuilt because we were dependent on Intel for everything from HR systems to finance systems to IT systems, compensation systems, you name it. Everything was built on Intel as the back end. The only thing that was unique to us was the technology and the customer-facing elements of the business, which obviously is important and the main part of the business. When we spun out in the spring of 2017 and finally separated ourselves from Intel, we really had to do everything new. We had new back office, new HR, new IT, new finance, all of it. We had a set of principles: we had to book, bill, ship, pay, and close. It was April 17th, I think, that we actually did it, and we were able to seamlessly get it done. But I will tell you, that's one of the hardest things I ever did, maintaining and running that business through a time where we literally cut over. It's like somebody taking the engine out of a car and putting a new engine in while you're still moving on the highway. It was a hard time, but it was a fun time. We did have some hiccups. There were some weird things. I remember this because it happened on Easter Sunday: there was a hacker that got one of the LinkedIn accounts, so we had to fix that issue. But those were small things in the grand scheme of it. I was really proud of the work that our teams did to get us there, and it made a huge difference.

One of the things you're pointing out about the work my team and I do at Microsoft is that we get an opportunity to be what Satya likes to call the frontier business group for the company. We're a small team relative to the rest of Microsoft, but because we're small, we can pivot and focus on some of the newer elements of technology that are presenting themselves and the opportunities that come up. Most things these days either feel AI-related or are AI-related. But even when I first joined the company several years ago, there were a couple of topic areas we were very focused on, before ChatGPT two years ago. For example, we saw autonomous transport as one category that we thought would be big. Fast forward to today, look at where we are with autonomous transport. I now regularly get into a Waymo and ride around. I live in San Francisco, so I can regularly get in a Waymo and ride around. It's a really good experience. Some of the things my team did is we invested in a number of players. We had a great investment in a company called Wayve, which is out of the UK, and they're building autonomous transport technology which will go OEM into a variety of different cars. I think we've got a real opportunity to change, from a societal perspective, how we move goods and how we move people. Remember in 2020 when the pandemic hit, how so many of us couldn't get goods and services, we couldn't move around, and there were shortages of drivers? If you think about what autonomous transport can do for us, it can change the paradigm remarkably. That's a category where my team has been able to make some investments and bring partners to work with us at Microsoft. We're doing some similar things on the healthcare side. Generative AI has really started to take the imagination in a lot of ways, but even before ChatGPT took everyone's imagination a couple of years ago, we started to make investments in companies like Paige AI, for example, which is using computer vision models and now a combination of computer vision together with large language models to do things like help pathologists diagnose cancer, and really diagnose undetectable cancers to most human vision. That's just one example of so many ways in which I think AI will massively change health outcomes for people around the world, and that's something personally I'm very excited about.

I think first of all, what's great about what we're experiencing right now with AI, and it was really driven by a lot of the excitement that came from generative AI, which became a part of our consciousness collectively in the last couple of years. But it's important to remind ourselves, as you're saying, that there have been many versions of AI around for a long time. We didn't just discover AI two years ago when ChatGPT was launched. What it did was make AI accessible to the average person in a way that was quite different than anything we'd ever seen before. What we did see in the last several years is an exponential increase in the capabilities of what these models could do and what they could bring to us. That's really what changed the game. But underneath it all, as you point out, there have been a lot of forms of AI. As I mentioned, in cybersecurity, we used versions of AI for a number of years. Microsoft, even pre-our relationship with OpenAI, has had many AI researchers and has been building models for a number of years. As you point out, there's going to be a variety of use cases. What I do think is important to note about the era we're in today is that AI is really something that I think will infuse our lives in almost everywhere we use technology. You mentioned the collaboration we did in bringing Steve Ballmer's vision here with the Dome to life. You can imagine a world where you'll be able to ask a question of the stadium: where's the nearest restaurant that serves this kind of meal? Where can I buy this particular jersey with this particular number? My car is parked in this part of the stadium, which gate should I go to? Can you tell me which direction to go? Think about the different experiences that one will have in the future as a result of generative AI. What I love about what we've done here is that we've laid the foundation for that to happen. Because of the way this stadium is architected, all those kinds of opportunities will be possible. If you think about it that way, it will completely change the experience. We're sitting in the Clippers' new home stadium. The NBA just released a new version of the NBA app, which now has the ability to send you more in-depth generative AI-related analysis around different players and games and moments in the games. Those are again just some examples of what's happening. But broadly speaking, generative AI is going to happen and will be part of our lives in many different ways.

That's a great question and probably the most important question for a lot of organizations. I think you need a couple of different elements that are important. One, which you're pointing out, is you need to make sure that your enterprise is ready. You need to have good standards around how you manage data and privacy. Those are some of the factors that matter a lot in terms of whether you can scale something out. Do you have the right security mechanisms and models in place? I think it's important to have the right principles around how you think about safety, security, and how you think about the output of the model, so responsible AI principles are really important. I think it's also very important to be clear about what use cases you're trying to solve. A lot of companies will get caught up in the challenge of having many ideas, but ultimately they're going to have to make some calls and really go and try to solve specific problems. At the end of the day, in the first phases of this, it's going to need to be leadership that drives it. I really do think that the CEOs and the CXOs of organizations are going to need to be the ones to mandate: this is what we're going to do, this is the path we're going to go down. If I just use Microsoft as an example, obviously we're providing AI to the world in many cases, and we use it in a lot of our own work at Microsoft. But so much of what's helped us be where we are today and leading so much of what's happening in this industry is because Satya himself was very personally involved in a lot of the decisions to go in the direction that we've gone, to really push, to really lead, and to lean into the partnership with OpenAI, which has really allowed us to be successful. I think that's a good example that leadership really has to embrace this. You can't just say, 'Hey, I'm going to ask people to go spin up a bunch of projects and hope that's going to drive success.' You've got to have those foundational underpinnings, you've got to know what use cases you want to go after, and then you've got to have the right leadership structure in place. You do need to allow people the ability to fail. You do need to be able to test and learn because you're not going to get everything right. But you have to have a long-term view of where you want to get to, and let the short-term pieces be worked out along that journey.

I think what's important to think about is that we have an opportunity, because we're still in the early stages of this, to shape the future that we want. I think that's very important. I'm a person who believes, I'm an AI pragmatist. I feel that we shouldn't sit back and be fearful of what might happen. I want to be part of defining what should happen. I think that's really been our approach as Microsoft: to say, okay, let's make sure that we have the right foundational principles around what should happen. We have responsible AI principles, we have AI adoption principles, and we're constantly testing those principles. We have mechanisms around that to make sure that we're protecting against some of the challenges that might be presented today, like misinformation or misuse of AI models. We want to make sure that our AI models are delivering useful outcomes as opposed to being misused for the wrong intentions. There will be things that will happen. We will have misinformation campaigns that will be powered by generative AI. With the ability to generate high-quality video and audio, you will have deepfakes that we'll have to deal with. But I'm confident that we'll be able to manage through a lot of those challenges because this technology has so much great promise in the long run. Like any new technology, we'll have to work through some of the ways it can be misused as well. But I'm one who personally believes that we have to define the future. We need to do it responsibly. We need to guide this technology in the direction that we want it to go because, whether you like it or not, we're living in the AI era now, and we will increasingly have it become a bigger part of our lives. I think it's a huge opportunity for us to take advantage of and do it the right way.

I think it's important. This change is going to have a profound impact. This is not just a technology shift that we're experiencing; this is something that's going to change the way we live our lives as humans. With that being the set of implications, I think it's super important that we engage with government. We have governments to help structure and manage how we live our lives and how we interact with one another. Rule of law, for example. Regulation can be difficult, but it also can be very good and important for us to have certain things that we expect to happen in our lives. I think that as technologists, we have to welcome interaction with the government. We have to welcome dialogue, we have to welcome exchange, and I think we have to welcome a certain amount of regulation to make sure that this all happens in the right way, that there are the right guardrails. It goes back to some of the things I was saying earlier. Responsible AI isn't something that just should happen at a company level; it needs to happen at a country level, it needs to happen at a global level. There will be tiers of people focused on how do we make sure that responsible AI happens, and that as we develop more capabilities, we have the right guardrails, the right partnerships, the right frameworks in place, and the right checks and balances, if you will, to get to the outcomes we want. Knowing that we won't get it all right every time, if we've got people in government who are educated and understand the technology, they'll be better able to adapt and learn and change as this adapts and learns and changes. I think that's really, really important. So we should want that, we should invite it in, we should look for opportunities to educate our colleagues in government who are not technical, because I think they're going to be a very, very important part of how this all plays out in the long run.

Indeed, indeed. Thank you very much. Our team, thank you. Thank you so much.