Ladies and gentlemen, please welcome Senior Vice President of CES and Membership for the Consumer Technology Association, Kinsey Fabrizio.
Thank you, and good afternoon. We are thrilled to have two exciting keynotes this afternoon, Qualcomm and Nasdaq. Most of us spend hours a day online and rely on our smartphones, tablets, laptops, and a host of other devices to stay connected and informed. We're also entering the age of generative AI, with new tools to discover customized information and content tailored for our preferences. Cristiano Amon is the President and CEO of Qualcomm and a member of the Qualcomm Board of Directors. In a career with Qualcomm spanning over 25 years, he has shaped the company's strategic direction, spearheading Qualcomm's 5G strategy and global rollout. Cristiano led the company's strategic expansion into new industries like automotive, and under his leadership, Qualcomm is leading the way in on-device generative AI, transforming how we engage with the technology around us.
He'll be joined in conversation today by Fox Business anchor and host of The Claman Countdown, a CES favorite, Liz Claman. Liz joined Fox Business when the network was launched in 2007. Today, Fox Business is the number one cable network in prime time. Liz has interviewed a wide range of industry leaders on her Emmy award-winning show, including interviews with every U.S. Treasury Secretary in recent history and the world's top business leaders, entrepreneurs, and newsmakers. Please join me in welcoming Cristiano and Liz to the stage!
Hi, everybody. Oh, this is... Wow! This is a big, big audience here. So, Cristiano, the pressure's on.
Yes, excited to be here.
First of all, welcome to all of you. Thank you so much for being here. We are gonna be talking obviously about all of the news that you have made so far, but I am going to pry open, or at least force you to pry open-
Okay.
Your playbook when it comes to everything that you're doing with generative AI and how it actually integrates naturally between people and the users in this room and everywhere else, potentially, and their devices. So let's start with this, 'cause I know it's a little cold and early to be talking baseball, but where, Cristiano, are we, in your opinion, in which inning, perhaps, which baseball inning, to really say this new generative AI, what, it's taking off like wildfire. Which inning are we in?
All right. Very good. It's a good question. I think we're in the first one, heading towards the second. I think there's an incredible amount of potential, a lot of things that's going to happen. Our job, and we have been very passionate about the role of AI in the devices that we carry, you know, whether it's your, your phone, your PC, the device that is not at the data center, the car, industrial devices. We've been talking about on-device AI before it was popular. I think our job was to create a computing engine that is gonna make that technology to run pervasively. And I think what we see right now, we are at a point that those capabilities in the processors and the semiconductors are becoming available.
The next step will be the development of the use cases and the applications, and it's good because we're starting to see those things coming to life, whether in phones, in computers, the Microsoft Copilot is a great example, and in the car as well.
I want you to bring us back to November of 2022, and I know a lot of you guys know what was happening back then, and that is that, Sam Altman and the OpenAI merry band of brothers and freakazoids up in Silicon Valley. It's okay, I'm from California, I can say that. They had unleashed the chatbot, ChatGPT, on the world, and it took off like absolute wildfire. I need to know how Qualcomm went into hyper speed at that moment to start packing in generative AI into the chips, when it was really quite a relatively new technology.
Yes, that's a good question, and maybe allow me to tell a little bit of story. I think the company DNA is... Because where we live, like, we're not in the data center today. Where we live today is-
Hold on one second. We're gonna have you use that microphone.
All right. Is it better now?
Yep, better.
Okay. So where we live is really, it's outside the data center, is in devices that we carry, the battery-powered devices. And the company DNA is really about how to do high-performance computing at a very low power. So about a decade ago, and we've understood that the next form of computing, you know, you have CPUs, you have graphical processor units, but you also gonna have the next form of computing, which is accelerated computing for AI, is gonna happen everywhere. And we've been busy developing the technology that allow you to do that without compromising battery life. For example, if you have something else doing on your phone, you still want the thing to last the whole day. So when ChatGPT came out with those large language models, and it was really incredible, one of the fastest adoption of any platform in history.
One of the things that it did, anybody in the streets know about GenAI and ChatGPT.
All of a sudden, we saw the intersection point when the company's interest in AI said: "We can run this on the devices, on this engine that Qualcomm has built." ... And that's why, for example, we have now in phones, in PCs, some of the platforms that has the fastest AI engine in all the devices, and that capability is available with you. You don't have. It complements the data center, and I think that's the opportunity for us.
Well, yeah, taking it from the cloud, which is where it is primarily, down to the device itself, and that brings us to all the users and potential users out there. We have a video from Qualcomm that we're gonna run, and what I want you to do is talk about how you plan to integrate and how you already are integrating that kind of thing right directly into the device chips.
Absolutely. So the interesting thing, Liz, is the AI is going to develop different from the cloud and the devices, and hopefully, that's what this video is gonna show. You're gonna have the AI available in the cloud, and I think you can do a ChatGPT query like you have been doing, and people have been experimenting with different things. But on the device, the AI is just running pervasively. So as you can see, some of those use cases, think of something as simple as you use your phone every day, and you text, like you and I can text each other. Everything that you type, and that data stays in your phone. Everything that you type could be a query for the AI. AI is thinking: Is there something for me? Is there something? Yes, there's something for me.
This is talking about we should, we should get together. We should have a meeting. This is the availability. Do you want to send a meeting invite? Or you're talking about some place that you went, the AI will bring the photos and say, "Do you wanna share that?" Or, or you're gonna say, "You know, I've seen this data," or, "I saw a video." The AI will say, "That was the video." You know, and, and it's going to be this assistant with you that's running pervasively. And I think another example of that is what we're doing with Microsoft. We built this engine when we announced, X Elite, has the fastest AI engine of any device that go into a laptop, and then now you have the ability to run those very large, language models, into the device.
We can do, you know, models that have billions of parameters running the device, and you get an instant response right there.
Okay, so if I'm hearing this correctly, and I think I am, it will look at your device, it'll see all the pictures, it'll see how you've written things in the past, how you've communicated with people, who you've dialed, et cetera, and it continues to generate and grow and learn?
Yes. What is now very unique, and I wanna tell you about the vision that we have and what is possible, and we're ready, that heading towards the second inning that's happening. The way we thought about computing platforms and how you interact, let's just take the smartphone as an example. You have the computing engine, that's kind of what we do. That's a Snapdragon. Then you have an OS, and then on top of that OS, you have applications. And then you, as the human, go in and touch one app and another app.
Yeah, there's a lot of touching. There's a lot of inputting passwords.
You start an app, and you're taking information from one app to the other app. Now think about something else happening. There is now this AI engine that is running and trying to do one job-
- predict your every movement and try to learn from you. It's the same thing when you take that same example to some of the things we're doing on the car. Let's say, for example, you leave your office, and what you do every time when you leave your office, maybe you call somebody from your family. So the car is gonna know, time to call somebody from your family. They want me to call for you because they already learned from you Mm-hmm.
I think that's where we're gonna see a change in user experience. You're gonna have this assistant that is gonna be there with you, is gonna learn from your behavior, and that's what's unique about doing AI on the devices. When you do a query on the cloud, you provide some information, the cloud doesn't know you and give you an answer. But when you provide the cloud and the device working together-
The device has real context. Where are you? Who are you? What is the relevance of that information? I think that's what makes it all this very exciting and transformative.
Exciting and a little nerve-wracking. I don't... You know, you-- There's a whole different conversation about the privacy aspect, et cetera, and where-
Stays in your phone.
Well, okay. That's. I think that is really certainly interesting, where you not cut out the cloud, but you certainly bring it down to the device. Let me just stick with this. You had mentioned Microsoft when you were just discussing that issue. Keeping it with the smartphone, which has been your bailiwick for so many decades, let's talk about OEMs. Are you in conversations with OEMs? Which ones are we working with? I mean, is it, is it Apple? Is it Samsung? Is it Huawei, Xiaomi?
So the AI capabilities, this really happen on our processor, and it's really happening on the Android ecosystem, where we provide the Snapdragon processing platform. You know, the chip that we actually got an award here to show, the 8 Gen 3, which is. It has one of the fastest engines for AI in mobile. Like, something like Stable Diffusion, which you provide a query-generated image, we can do in 0.6 seconds, as an example. It used to take 10, 15 seconds on a PC. We've seen a lot of the phone OEMs preparing the use cases. So a couple things to watch. As this device is launching, we have upcoming device from Samsung, and we have been, there's been a lot of conversations between us and Samsung, us and Google.
You see Google talking about some of those new use cases Mm-hmm.
- on Android. So I think there's gonna be the beginning of having those incredible AI use cases coming and those new devices launching in 2024. We also saw other customers, companies like in China, Xiaomi and Vivo and Oppo, et cetera, already announcing AI use cases, and we're just at the beginning. We see cases with photography, we see with texting, we see with translation, but there's gonna be much more.
Okay, so let me just keep it on China for a second. How limiting is it for the U.S. government's regulations that they put into place for some specific chips when it comes to what you guys are trying to explode out into the world?
Right now, we're not impacted. Like, I said, we're not in the data center. I think we have been really focused on phones, PCs, car, industrial, and, to this point, I think we have a vibrant, business globally, on those products.
Let's get to laptops. How does Qualcomm specifically envision generative AI work? Forget envision, what are you doing right this second with laptops?
One thing that is really exciting for Qualcomm, and this one I'm gonna say GenAI is actually a tailwind, right? We put ourselves in a mission, in a partnership with Microsoft, to build next generation PCs. Think about the conversions between mobile and PC Mm-hmm.
Think about what users want of their PC. They want more mobility; they want a device that is faster, it has more battery life. And that was the mission on Snapdragon X Elite, including us developing our own CPU, which is the Oryon CPU. And we were on a mission to build the fastest processor for any PC. That was the role of X Elite, and we delivered. We announced that last year at a tech summit, and it was. The performance was incredible, actually exceeded our expectations, because on the CPU alone, it was faster than anything out there, that of any processor for any laptop. But the good thing is, that happened at the same time that GenAI is coming to the devices, and I think that's an incredible tailwind, because if you think about the Copilot, we can run the Copilot into the device.
The device has an incredible performance separate on AI. We talk about 75 TOPS, tera operations per second of AI computation, and you can run those things across every app. So I think as Windows change to an AI and a Copilot experience, X Elite is the platform that is gonna take that next generation Windows experience to consumers, and that's a growth opportunity for Qualcomm.
The challenges that your engineers face. Okay, I've been in San Diego at Qualcomm's headquarters, and you guys have lots of really smart people fanning out. But bringing it back to when ChatGPT came out, I mean, what were the conversations about, "Guys, we've got to get on this. We've got to hammer more transistors onto each tiny silicon chip." What challenges did they face in getting up to hyperspeed?
Yes. It was a combination of two things. The first thing. The first reaction was: Wow, finally, this engine that we built, we have an incredible application for it, and we were in the right track Mm-hmm.
The second thing is-
Were you ahead?
Yes, because I think it was very innovative to build that type of computation to a device that was battery-powered, not in the cloud. The second reaction is: How do we get those models ported into the device as fast as we can possibly can? It has been, the last quarter, a mad rush. Just to give you a perspective, you talk about ChatGPT, and there are a number of models, also open source models. There's Llama from Meta. Right now, we have over 40 different models ported into the Snapdragon, and I think that's been incredible amount of work over the past three months to get this ready, so applications can come to life in new products launching this year.
You referenced cars. We just came from the West Hall. I don't know if you guys have been over there, but it is—it's basically the Detroit Auto Show on steroids. The CES show has become the auto show. It's, it's, it's a de facto auto show at this point, and it's fascinating to see, 'cause I can remember being here many years ago. This is my, I don't know, seventeenth year, and Alan Mulally of Ford was the first auto CEO to be invited, and he stood outside. They had a car outside, it wasn't even inside, and they were talking about Sync, which back then was sort of that first get in the car, "Oh, it knows it's your phone," and here we go. We've come so far from there.
Let's talk about the digital cockpit, and when you think about consumer electronics, the car is the most expensive one most of us have.
Yes. Look, I love talking about cars, and I think CES, to a great extent, it's a car show, among other things. And I can spend a lot of time talking about it because there's fundamentally one thing that is different now, and I think that's the foundation of many of the things we're doing with the Snapdragon Digital Cockpit and some of the rapid success we had in automotive. The car is a new computing space. That's probably the best way I can describe it. You have your phone, and the phone, you do a lot of things. It's an inseparable device. You have your laptop. Laptops didn't go anywhere. They continue to exist. They will continue to exist. Even you have some cases, you have your desktop.
Desktop's still there, didn't get replaced by the laptop. But now you also have the car. The car is a computing platform. If you look what people are doing in their cars and why we see those immersive screens that we're enabling in Digital Cockpit, you're getting entertained in your car, you have productivity in your car, you're communicating with people in the car, so it became a new computing space. And so much so that, like, you see many companies, one of their applications and services going into the, "I need to be in the Windows PC, I need to be in the phone," you see they're going to the car. Like, a great example, we had a great partnership, we have a number of examples at the show with Salesforce, that Salesforce embedded into the Snapdragon Digital Cockpit so you can do CRM from the car.
You can schedule different things from the car. We have insurance companies wanting to embed their services into the car. You have distribution of television content, so gaming. So the car is a new computing space, and that is a revolution that's created a huge opportunity for the car companies and Qualcomm role to provide that digital engine.
... I have never seen the car people here at CES so excited. We just talked to Dimitris Psillakis of Mercedes. He was vibrating. I mean, he was so thrilled, and I know it wasn't 'cause he was just getting interviewed by me, but I thought it was really fascinating to see how the sort of levitation of all that is going on in the car now technologically, but the generative AI aspect of it takes it to a whole new level, does it not?
Absolutely. Look, we talk about it in some of the videos I think we show, talk about the GenA I, providing some experience in the phone. In the car, it's a little different.
One thing that you look at those large language models, I personally think the car is an incredible use case of how we can use voice, because when you are behind the wheel, voice and visualization become very, very important ways for you to communicate with the car. And what is GenA I doing with the car right now is actually creating an opportunity that you actually have a conversation with your car. So we have a demo, you know, the concept car, that you were there in San Diego when we first unveiled that car.
We have a demo, and the demo shows with something that we call, you know, the intelligence service manual in the car. So a light blink up, you ask the car: "What is this light? Why is it blinking?" So the car, you know, knowing, think about a ChatGPT-like, know the information from the car maker, will tell you, "This is what it is. This is the reason this light is on. This is what's causing it. Here's how you're gonna repair it." And then will even ask you: "Do you want me to schedule a repair for you?" And using Salesforce, integrating with some of the dealership will say, "Those are the slots available." The same thing, you can ask your car to plan things for you. It's like a large language model.
It's like doing a chat query, but you're now chatting your car. And we're gonna get to the point very, very soon that things that used to be one of my favorite shows in the eighties, Knight Rider, that you can talk to your car.
I remember.
I think that's possible right now, and it's just as those things continue to develop.
Yeah, it's, it's definitely happening. In fact, one of the things that we saw with Mercedes was it has synced the music to the way the car drives. So if you have your favorite song on, let's just say, the Eagles, Life in the Fast Lane, it will speed up a little bit when the music's faster, and then slow down. Of course, it must, it must know that there's somebody in front of you, so it doesn't just hit it. But, you know, it's, it's really the future is now, and it's happening at light speed, which I think is truly fascinating. I need to know what scares you. What scares you about all of this development? There's an arms race. People are running over each other to try and beat the other guy in developing this stuff. What's the cost that you worry about?
Look, let me maybe give you a Qualcomm answer. For us, this is exciting because I will say, in this space, we are the underdog, right? This is all growth opportunity for Qualcomm. I think you said it correctly. I think we have been well-known for what we have done and continue to do in the mobile industry, but we're really changing the company from a wireless communication company to a connected processor and artificial intelligence company. So for us, this is all new growth opportunities, and I think it's starting to show based on what we inroads we made in automotive, the inroads we're making in the PC. And what is good for Qualcomm is our mobile DNA, because our mobile DNA, the mobile creates a very difficult engineering environment. If you're designing chips, the mobile give you a couple constraints.
The size is very small, the battery is small, it has to last a whole day, it can't be hot because you're gonna touch your face, and you need to increase generation of a generation, the computing power. When we take this to other industries, I think it give us an advantage, so we're actually excited about the opportunity.
Let's talk about how Qualcomm envisions that relationship between AI and augmented and virtual reality. The headsets are a big, big topic of conversation and a big trend here on the floor at CES. What are the possibilities? We talked to HTC VIVE, and they're already moving into big enterprise usage of their headset, and I just thought it was really interesting. Firefighters are training with them. You can feel the heat. They have sensors on you, so you can be pushed back by fire when you're training, versus actually having to burn down a whole building. If you know anything about firefighting, they have to subject themselves to that. That's a $300,000 set that they have to build and then burn down. What do you envision there, both for the consumer and for the enterprise opportunity?
Yes. Spatial computing is going to be another computing platform that is going to eventually get a lot of scale. And those devices that we have been helping build, they're becoming more affordable, they're becoming smaller, they're becoming more capable. And the simple thing, I think we use this way to describe it, which is: How do you merge the physical world and the virtual world together? And it's going to have a number of applications. For example, I like what you said about enterprise. We're starting to see now, you started with consumers. We have great partner into Meta with their, the Quest devices. You see a lot of consumer, consumer gaming, but now you see some of the use cases of enterprise, how you train people. How do you enable you to be fully immersed?...
into that digital environment, how are you going to evolve communications? The communication that we have been—you know, we started talking on the phone like this, then we're texting each other, then we learned to hold and have a video call. It's gonna be a hologram, as you think about into the future, because you're just gonna use those augmented reality glasses and virtual and mixed reality. So that's a huge potential. We have been partnering with virtually every company right now. You see what Meta is doing. Google, Samsung is coming next. Sony announced a device. I think Lenovo is building enterprise devices for some of the enterprise customers. With HTC VIVE, you saw some of the use cases. I think that's a great opportunity. It's just a matter of time.
Is your phone ringing off the hook with all of these guys calling you up and saying, "Quick, make a chip for this?
We're very fortunate. I think I have a number of partners. I'll tell you right now, I'm so glad... The phone was off the hook when we had the supply shortage. Right now,
Uh-huh.
I'm so glad that that's behind us, and we're just thinking about.
Totally behind you? You don't see any-
Oh, 100%.
-shortage anymore.
100% behind us right now. I've been careful, I think, when I say, but, you know, the phone market right now, I think it's getting back to normalization, and, we're excited about the growth that's starting to show in all those different, vectors.
What you just articulated to me brings up this question as we finish up, Cristiano. Will AI eventually create a way, whether it's through holograms or something entirely different that we haven't even envisioned in this room yet, will it eventually kill the smartphone?
I don't think so. I think what, what is happening, and that's a, that's a big topic of conversation. I wish we had more time.
We do.
But what-
We got 44 seconds.
No, we keep going.
What AI is doing is really merging the smartphone in the cloud, and it's changing how you think about the computing platform. I'll give you an example. Many people use different ways to describe it. They talk about disaggregation, but I think the way to describe it is, let's say that the AI know that based on your calendar, based on the fact that you drop your car at a service and you need to go somewhere, the AI know that it needs to call an Uber for you. And the AI will say, "Liz, do you want me to call an Uber for you?
I already know from this place at that time to that other time." You're gonna say, "Yes." Well, at that time, the AI could do one of two things: It could go to the Uber app in your phone and call an Uber for you, or it could just, just go to Uber straight into the cloud and call an Uber for you. It doesn't matter. What's gonna happen now, the phone is gonna have a different role that not only it's gonna be like what it used to do, which is the phone is good for all the apps that you have on your phone, but also the phone's gonna be communicating with the apps in the cloud. It's going to have a big augmentation. The phone's not going anywhere, but it's gonna get enhanced by AI.
It's gonna be get enhanced by the glasses you're gonna use it, the same way it's got enhanced by the watch that you use it, and it's gonna get enhanced by what you're gonna be doing on the PC.
Cristiano Amon of Qualcomm, we thank you. We wanted to pull all this out of you, and you've done it in a very short period of time for an audience that is incredibly motivated to hear exactly what you foresee in the future for Qualcomm and everything it can do for Gen AI, AI. Thank you so much.
Thank you, Liz. Pleasure.
Terrific stuff! Thank you so much.