On this episode of The Six Five – On The Road, hosts Daniel Newman and Patrick Moorhead welcome Microsoft’s Eric Boyd, CVP, Azure AI Platform, and Ross Kennedy, VP, Digital Natives, for a conversation about Microsoft’s AI approach and leadership from Microsoft Ignite 2023.
Their discussion covers:
- What is driving the AI wave and how Microsoft is helping businesses adopt AI
- How Microsoft working with OpenAI to bring the latest generative AI innovations to Azure
- Key Announcements from Ignite on Azure OpenAI
- How startups and born-in-the-cloud companies are innovating with AI
- A look into how customers can keep up with the rapid pace of innovation
Watch the video here:
Or Listen to the full audio here:
Disclaimer: The Six Five webcast is for information and entertainment purposes only. Over the course of this webcast, we may talk about companies that are publicly traded, and we may even reference that fact and their equity share price, but please do not take anything that we say as a recommendation about what you should do with your investment dollars. We are not investment advisors, and we ask that you do not treat us as such.
Patrick Moorhead: The Six Five On the Road at Ignite 2023. And Dan, we are talking about some of our favorite, favorite, most favoritest topics here, which of course is AI, which is data, and what customers are doing with all of this magic. How are you, my friend?
Daniel Newman: Yeah, it’s good to be here. Microsoft Ignite is the big Microsoft Marquee event each and every year. And for The Six Five, we love being on the ground, whether that’s physically or virtually, to talk about what’s going on. It’s been a huge year with all of the momentum for not just AI, but really generative AI, which is changing the conversations, many businesses, rethinking how they’re going to go to market, thinking about how they’re going to staff, thinking about how they’re going to go through their workflows and how they drive future versions of their go-to-market strategies. It’s going to come down to differentiating with technology, differentiating with AI, Pat. And Microsoft has been at the forefront of this discussion. You and I on the record many times now, talking about how well Microsoft has done, getting out in front, having a great story to tell, and, of course, helping the market understand how this is a good business, not just for Microsoft, but for the customers that Microsoft serves.
Patrick Moorhead: That’s right. Customers want to know what are the advantages, what are the disadvantages of folks, and I think they’re looking for real world examples, maybe somebody in their own industry. But without further ado, let me introduce Eric and Ross. First time on The Six Five. Welcome.
Ross Kennedy: Thanks for having us.
Eric Boyd: Thanks for having us.
Patrick Moorhead: Great to be here. And like Dan said, whether it’s on the ground virtually, I know that I just came off a video version of some Ignite activities, but it’s great to be here. Thanks for coming on the show.
Eric Boyd: Of course.
Ross Kennedy: Very excited for Ignite.
Daniel Newman: It’s great to have you both. Eric and Ross. Let’s start off with you, Eric. There is a lot to unpack. Clearly it’s going to be a big week, lots of announcements, whether it’s talking to the business, talking to the developer, and more and more that’s becoming the same thing. The utilization and opportunities for AI are compelling. Microsoft’s been innovating rapidly. We were on the ground, both Patrick and myself, all the way back in the very beginning of your open AI announcements. We’ve been with you through a number of events, most recently in New York, and now here at Ignite. I mean, give us the backgrounder here. What’s kind of the ethos at Microsoft driving this wave, and then, of course, the approach that you’re taking to helping businesses adopt AI?
Eric Boyd: Yeah, sure. I mean, obviously, we’ve just been kind of astounded by the interest in what’s happened over the last 12 to 18 months and the amount of customer interest and every business out there really working to figure out how can they participate, and what are the things they need to do. And so Microsoft has really been focused on working with our customers in a couple of different areas. First off, we’re building AI directly into our products. And so you saw that first with GitHub copilot, which we announced, gosh, almost two years ago, now, really helping developers become much more productive as they’re building code and developing and all of that. And then we’ve really furthered that across the full product suite of Microsoft with really the highlight being M365 copilot, our big copilot for business users to be able to really understand all of the things that they’re doing from their email to their documents to their calendars, to just be able to have conversations with an agent that’s helping make you more productive on that. So the integration is a huge thing, but the other side is we know there are a lot of applications where customers are trying to build their own applications, and so we’ve created a full suite of tools making it really simple for customers to just go to the Azure AI studio and just integrate all of these different pieces together to build really compelling applications. The third thing is we really know that every company out there is really focused on making sure they’re doing this in a responsible way. They want to make sure that the AI reflects their brand, it reflects what they’re trying to show to their users, and they want to make sure that they’re doing it responsibly. And so we brought a whole bunch of tools and integrated that directly into all of our services to make that simple for our customers. So those are the main things that we’re focused on, but yeah, tremendous interest and excitement from customers really all across the spectrum.
Ross Kennedy: Yeah, it’s great points, Eric, actually, and just to echo that, the interesting thing is, normally when you see a technology curve, whether it’s Web3 or blockchain, there’s always early and late adopters. We’re not really seeing that with generative AI. We’re seeing across the entire spectrum, companies that are in all different stages of their own growth and their AI journey that are adopting Azure on OpenAI from NBA to Volkswagen to Providence Healthcare, to those that are truly born in the cloud, companies like AltaML, Commerce.AI, Motto, and MakeMyTrip. It’s an exciting time to be in technology.
Patrick Moorhead: It is exciting, and it’s funny, even though we know all the dates and how far AI goes back, and the algorithms developed in the 1960s, but it’s funny, the common person test was, hey, when did generative AI really take off? And it really had a lot to do with your relationship with OpenAI and that first event that Dan and I attended at your headquarters, and that was about what you’re doing with OpenAI, how you were approaching it both on the consumer and the enterprise side. And I get a lot of questions on how does Microsoft work with OpenAI additive to bring some of these latest innovations to Azure AI. And maybe, Eric, you can kick that off.
Eric Boyd: Yeah, happy to. So I mean, we’ve obviously had a long partnership with OpenAI, I mean, long and AI terms, decades in AI terms, if you know the speed of this space. A really tight relationship working with them to figure out where do we see this space going, and how do we help get there together? And so of course, OpenAI runs all of their systems on Azure, and we partnered very closely with them to make sure we understood what is it really going to take to build these cutting edge models, and how are we going to bring them to customers. OpenAI had their Dev Day a couple of weeks ago, and Sutskever on stage talked about how we’ve really customized our infrastructure from the data center all the way down to the rack, the whole stack, all the way through to make sure it’s fully optimized for AI. And that’s led to just some really amazing progress in terms of how everything can work together. And then now, we’re starting to see that show up in all of these announcements and models and things. And so at their Dev Day, they talked about a bunch of new models including GPT-4 Turbo and all the advances that’s going to bring. And so it’s really that partnership of making sure we’re working closely together on where do we want the models to go, where do we think the technology’s going, and how do we have the hardware and systems and infrastructure really to make all of that possible that all combines to really create this magic that we’re seeing in the industry right now?
Patrick Moorhead: Yeah, it is amazing the symbiotic relationships and what you’re able to crank out. And quite frankly, a lot of people forget that OpenAI runs on Azure. And I know that means a lot to me, because there’s a lot of different infrastructure that you can run on with different capabilities, and the fact that it’s running on Azure I always thought was impressive.
Eric Boyd: And just to build on that, we hear other companies coming to us and saying they need that same infrastructure, because they’re looking for how can they build the latest and greatest models? And virtually everyone who’s trying to build cutting edge models is coming to talk to us, because we’ve really thought through this infrastructure in a really deep way.
Daniel Newman: Yeah, it’s really evident, too, that the fast innovation for generative AI is going to happen in the cloud. And Pat and I are not deniers of the idea that hybrid and multi will be a very important fabric for most enterprises, but we’ve just realizing even companies with really long legacy of prem that have large customer bases, and I’ll let everybody out there guess who we’re talking about, but they’re pretty much saying, you want our best features, our best generative capabilities, it’s going to be built in the cloud. And so you find yourself in a really good position right now, Microsoft, to help companies with this. But hey, we’re here for Ignite. So while we definitely love talking about what you’re doing around AI, and I think that’s been really a commitment from Sutskever through the entire organization from Microsoft to lead with AI this year. But as you’re focused on the Azure OpenAI opportunity, what are some of the key announcements, Eric, that you have there?
Eric Boyd: Yeah, a lot coming. So, obviously, we’re following up the announcements from OpenAI’s Dev Day, all of the things that they talked about, we’re working to bring them into Azure. And so the GPT-4 turbo will be in public preview, the GPT-35 turbo with 16K that’s been updated in November, that’ll be in GA, the GPT-4 turbo with vision that’ll be in preview. And so just getting all of these new models and capabilities into customer’s hands, so that they can go and build interesting applications with it. But on top of that, there’s a whole bunch of work that we’ve been doing to make it simpler for users to develop these applications, these gen AI applications. We’ve seen such rapid adoption because they’re not that complicated to build. You have to sort of structure a few things together, but even those few things can be made simpler and people get sort of a little overwhelmed with terms like, oh, I need a vector database or a vector index, and I got to put all of that together. And so the Azure AI Studio is the unified studio that brings all of the pieces together from your data and how you manage that and pulling data in from fabric to integrating that into the latest models, to orchestrating it and really giving you the full lifecycle management that you need to develop those applications.
And I talked about vector search. So Azure AI search brings the latest of vector search combined with sort of the traditional relevance and ranking techniques that we’ve been evolving for years and bringing into a single search place. And the combined results are actually much more powerful than either using traditional search mechanisms or vector search alone. You combine both of them and have a semantic ranker on top that really brings you highly relevant and customized experiences from that. And then as users are developing these applications, they often need to sort of manage this full life cycle of the model starts, and now I need to refine the prompt or think about different ways that I want to test it out and make sure that it’s working. And so prompt flow is a tool that we’re bringing to make it really simple for users to manage that full application stack. So, a whole bunch of things that we’re doing to try and make it easy for developers to develop these applications. On top of that, we announced the customer copyright commitment, and this is now going to include the Azure OpenAI service, so customers can know that if they’re using our service that they’re going to be protected against the model saying things that could get them in trouble. We’re also bringing a number of models as a service onto our platform, making it just really easy for customers to consume. So just a whole lot of things happening really across the full spectrum, with the key focus being how do we get the latest and greatest technology in our developers’ hands, and how do we make it really simple for them to create compelling applications that are going to deliver value to their enterprises?
Ross Kennedy: Yeah, it’s an incredibly exciting time here at Ignite. As Eric said, we’re really seeing businesses adopt generated AI, not just for those customer facing applications, but also to boost productivity of their teams. A couple of great examples include Siemens and Perplexity AI. Siemens is improving manufacturing operations, safety, and reliability for its customers. And they do this by connecting shop floor workers using Teams. They use predictive supply chain management to ensure on-time delivery, and they have preventative product detection through industrial AI. Perplexity AI, I can’t leave it to the side, very exciting startup, focused on conversational search. They were able to speed time to market in their AI engine using copilots. So it’s an amazing time to be here.
Patrick Moorhead: Yeah, I do love hearing about real world customers and real world case studies, and I’m sure we’re going to hear a lot more of those at Ignite. So Ross, I want to talk to you about what’s… In fact, this name or descriptors in your title about digital natives. Typically, the first movers, it could be startups, they could be born in the cloud versus let’s say a brownfield company that’s 150 years old, which is all interesting as well. Can you talk about that tip of the spear and how they’re innovating with AI?
Ross Kennedy: Yeah, yeah, that’s a great question. Just like that Perplexity company that I mentioned earlier, we have thousands of these startups and even at scale born in the cloud companies across the globe using Azure OpenAI and related services. Customers are choosing Azure, typically, because of scale and security, but most importantly because of the speed of innovation that they’re able to deliver on the platform. We have customers driving innovation across all verticals. There’s financial services, retail, healthcare, travel. Some of the exciting examples that we have. AltaML, for example, are building a powerful AI tool to help firefighters in the government of Canada better predict and sustainably manage risks of wildfires, creating the potential to save millions in annual operating costs. Two other examples, Pangaea Data as well as BeeKeeperAI, are using the power of generative AI to improve patient outcomes and drive efficiencies in healthcare systems. Another one would be Commerce.AI. They’re leveraging generative AI to drive up to 30% productivity gains in support centers. So you’re finding that it’s not one example, but it’s many examples across all industries. And those tip of the spear are really trying to find unique and compelling ways to use generative AI to drive innovation and disrupt the industries themselves.
Daniel Newman: Ross, I’ve been reading about some of the things that Microsoft’s doing with Y Combinator. It’s been offering resources, virtual GPUs, free tier services. I recently had a call with another venture arm of a big tech company, and it’s not just about the money anymore in these types of relationships. It’s also about how do we enable, how are we an enabler? How does being part of our ecosystem give you advantages? Whether that’s the people that might connect to the network, it could be the access to data, of course, technology systems and mentorship. And I think after 2021, a lot of people realize after plenty of features got taken to market, and now we’ve got torn back down that we’re going to do it differently this time. I mean, talk a little bit about Microsoft’s thought process around enabling and partnering with AI startups and how these startups could work and engage more closely with Microsoft.
Ross Kennedy: Absolutely. The Y Combinator announcement is obviously very, very exciting. You think about it, access to high-end GPU clusters is key to those companies trying to drive AI-based disruption. And if you think about the use cases by which you would develop on generative AI, they’re resource intensive and often very expensive. So by making AI infrastructure more available to startups, we believe we can help them accelerate on that journey. But even sort of stepping back and thinking about how Microsoft helps those startups get there, there’s probably two programs that come front to mind. The first one is a program called Founders Hub. This is a program by which companies can register, and you’ll gain access to everything from credits to education to some of the latest information on releases, as well as the access to some of those emerging technologies. Another program is called ISV Success. So for companies who have a developed solution, whether it be on generative AI or not, this enables them to put their solutions out onto our generally available marketplace to reach thousands upon thousands of customers they otherwise might not have, allowing them to turn their energies to doing what they do best, which is to continue to innovate and develop. Essentially, no matter where you are in your journey within generative AI, Microsoft has a program there to help you and to help you be successful finding an audience for it.
Patrick Moorhead: This is a great conversation. And by the way, on the Y Combinator discussion, I mean I see funny memes all over social media. We will work for GPU Access.
Daniel Newman: Won’t we all.
Patrick Moorhead: Some really fun ones, but it is just fun to see out there, and it’s great to see that Azure has the capability to provide that, because quite frankly, 10 years ago when Microsoft was first getting into the cloud, they weren’t the first movers who you would talk about. They were very established, very large customers. Typically, had their cadre of Windows server on prem. And I just find that fascinating. It shows to me the different lens to which we’ve looked at Azure over the time, in particular during this generative AI moment, how different that people are looking at Azure. So it’s been a great conversation. I wanted to hit both of you with kind of a final question here. We’ve seen just a blizzard of announcements, I would say, since the beginning of last year. And it’s one thing for those in the tech industry who eat and devoured this, and I would say particularly easy for analysts like Dan and I, because this is our full-time job to track it, measure it, but people have real jobs. They run businesses, they run IT, they run a department, or something like that. How on earth do they keep up with this pace? And by the way, I think we also risk turning people off that it’s like, well, I’m going to wait. This is going way too fast. You all remember on the playgrounds, some of you might be old enough, where you get out, and that thing is spinning around at light speed, and you had to time the right time to get onto this thing. Some people might be like, I’m not getting near that. That looks too dangerous. Anyways, very long question here, but maybe Eric, we’ll start off with you.
Eric Boyd: Yeah, I mean, it’s a question I get a lot from a lot of customers, and I love the merry-go-round analogy. It’s been a while since, I think. Though they’re safety hazards, I don’t see those of lately.
Patrick Moorhead: I think they banned them. I don’t know if those exist anymore.
Eric Boyd: But as you’re looking to jump on, I think the, hey, when is an important question. If you look at how Microsoft has approached this, we’re all in, every business is jumping on. I mean, it really goes back to when we first saw GPT-4, such a challenge in every business unit to come up with your plan for how this is going to transform your business. And that’s led to this crazy string of announcements that we’ve had and just this re-potting of the company around how can we take advantage of this? And so I think companies around the world are kind of seeing, hey, this is really important, and I need to get out on this too. So how do you keep up? How do you learn? I think the key thing is to get on the merry-go-round. And so figure out what’s the first application and start small. Start with something that you can try out and learn from and just really learn by doing, because that’s what everybody’s doing in this space. And just know that it’s going to evolve, and this stuff is changing. But if you’re in the game, and you start to figure out how this can work, can you leverage some of Microsoft’s existing copilots? Can you build a small application of your own, and then you just learn and grow from that? And the companies I’m seeing most successful, that’s the approach that they’re taking is start with something and then really grow out from there.
Ross Kennedy: Yeah, I have to echo what Eric said. So I’ve had the pleasure now of meeting more than 300 of our digital native partners over the last year. And the most recurring theme over and over again is just get started. You have to get onto that merry-go. I personally wouldn’t. I’m terrified of merry-gos, but for the AI merry-go, I’m all in. But it’s absolutely right. Get in, and you’ll find that as you continue to develop, as you continue to find your footing, you’ll end up discovering what you need to know along the process. It’s easier to get yourself dug in than it is to try and read about it.
Daniel Newman: Well, Ross, I can be 301, and I can tell you we’re building an intelligence data platform on top of Azure OpenAI services. So we’re actually using, we don’t only talk about it as analysts. I always say it’s very indicative of the technology we believe in by figuring out the real stuff that we use as analysts every single day. I won’t go through the whole stack right now, but Pat alluded to earlier some of the things he’s doing. So as you can see, definitely a fan. Appreciate you both taking the time and spending a little bit of time with us here. We know Ignite is a really, really busy time and a ton of exciting announcements here. So thank you so much for sharing. Congratulations on the progress so far. And you can be sure that both Patrick, myself as analysts and as the Six Five will be keeping an eye on all the things you’re doing over there at Microsoft.
Ross Kennedy: Great. Thanks so much for having us.
Daniel Newman: All right, everybody hit that subscribe button. Join us for all of our Six Five episodes. Look for Pat and I to cover Ignite more deeply on our weekly show. And, of course, stay with us for all the content from all the leaders across the technology and industry space that come and join us here on The Six Five, every single week, year in and year out. But for this episode, for this show, from Patrick and myself, time to say goodbye. We’ll see you all later.