SoundHound AI Co-Founder and CEO Keyvan Mohajer sits down in-studio with Yahoo Finance Live to discuss the company's artificial intelligence-powered customer service products and the broader opportunities AI adoption is creating.
"The reason we are not talking to devices yet because is the technology wasn't ready. Now, because of generative AI it is ready and fortunately for product creators, a microphone is all they need — a very inexpensive. small microphone is all they need to enable their devices with the most natural interface," Mohajer explains. "We are enabling cars and TVs... millions of devices, and we are powering customer service, we are in 10,000 locations with 100,000 in our pipeline..."
Joining us now, Co-Founder and CEO of SoundHound-- that is Keyvan Mohajer. Thank you so much for being here. Appreciate it.
So if you look at what we think is going to come next is the wave of experiences that people are going to use generative AI for to create a lot of value. So the analogy is iOS and Android were the big bang. And then for more than 10 years, apps, and apps, and apps. And some of those apps were new experiences for the first time created billions of dollars of value. And we expect the same thing will happen with generative AI. So years of value creation and experiences, and that's what SoundHound does.
- And, Keyvan, me ask you just about that Nvidia stake. I'm just interested to hear more about how that came about, Keyvan, and what do you plan now to do with that money. How does SoundHound put that to work?
KEYVAN MOHAJER: Yeah, so we've known Nvidia for a number of years. And they are a great company. We respect them a lot. And they specialize in-- they create the infrastructure for AI, and SoundHound puts that infrastructure to good use. So the synergy is very clear.
And the amount of investment is not as big-- we are a public company and well-capitalized. But what it does for us is two things-- validation and alliance. And we've attracted a number of strategic investors like Hyundai, and Oracle, and Vizio, and Samsung have invested in us. And it's validation because they bring their own experts to look at our technology and our vision.
And when they take the equity, it checks the box. And alliance, because it brings us closer together, and we are able to turn some of those into really big customers, like Hyundai.
- And, Keyvan, another question, though-- these reports, I saw, Keyvan, and you saw them-- options volume and sound in your company kind of surged here in the days leading up to Nvidia's filing, announcing that stake. That can be the thing, Keyvan, kind of get regulators sometimes do get interested in. What can you tell us about that?
KEYVAN MOHAJER: Speculation. And I think it's wrong. Well, first of all, who knows? But if you look at-- they looked at some companies that Nvidia invested in, and some volume went up before the announcement. But if you look at all the other AI companies, like Big Bear, their volume went up too. So I think it was unrelated to Nvidia.
JULIE HYMAN: Yeah. But to your point, there has been this-- we were just talking about it with Josh Schafer a few moments ago-- this enthusiasm, maybe even this mania that people just are falling over themselves to invest in companies like yours. Now, you say that this is bigger than the internet, that this is a huge opportunity. So is something like SoundHound undervalued, then, at these levels, if that's the case?
KEYVAN MOHAJER: Well, we-- obviously, we think long-term. And I'll tell you what we are doing, and you can judge. So we made two predictions, and we are focusing on those. The first one is that AI customer service will be as necessary to every business as Wi-Fi and electricity.
So you create a new business, you sign up for an internet, you sign up for electricity, you sign up for AI customer service. And we provide AI customer service for businesses. The second prediction is that we prefer voice as an interface to devices, to physical products.
So I have a three-year-old daughter who knows how to speak. And she's going around the house and talking to everything. And not everything is listening to her, but she knows how to speak. I still have to teach her how to type, and use a mouse, and everything, but speaking is a natural way.
And the reason we are not talking to devices yet is because the technology wasn't ready. Now, because of generative AI, it is ready. And fortunately for product creators, a microphone is all they need. A very inexpensive, small microphone is all they need to enable their devices with the most natural interface.
So we are focusing on that too. So we are enabling cars, and TVs, and IoT devices-- there are millions of devices, and you are powering customer service. We are in 10,000 locations with 100,000 in our pipeline, with 30 million businesses that can be enabled-- that's a $100 billion opportunity.
JULIE HYMAN: It's an opportunity that a lot of others are chasing also, right? You do have competitors in the market who are working on this same problem. What makes SoundHound different? Is it the data set that you're drawing from? Or what exactly is it?
KEYVAN MOHAJER: So there's the big tech, and there's the smaller newcomers, right? So we've been in this for 20 years-- it's a vision that we came up with in the Stanford University dorm room, and we've been pursuing it for a long time. So when the generative AI revolution happened, the big tech had to go and build their own.
They're not going to go and use an OpenAI API, right? And that's going to take them a couple of years. So they're going to give us time to go faster. And the smaller players-- generative AI is not that easy to make it work, right? So it's not a plug and play. You have to do a ton of software engineering to reduce hallucination and to make them do the right thing, to go from a demo to a live product. And we are really good at that. And that gives us the advantage.
- Keyvan, I'm interested-- does your product work better-- and by "better," I mean does it respond more quickly and more accurately than the voice engines a lot of our viewers would be familiar with from big tech, from Google, and Amazon, and Apple?
KEYVAN MOHAJER: Yeah. We actually we have been, for a long time, known to be the AI that handles more complex queries. Like, you can ask, show me Asian restaurants, but not Chinese, not Japanese, maybe you had Chinese food yesterday, that are open after 9:00 PM on Thursdays and have more than three stars and are within walking distance of the Space Needle, right?
So we can actually handle queries like that. And now that we've integrated generative AI, you can actually have an endless conversation with it-- ask for recommendation, and opinions, and gift ideas, and go back and forth between them. And we were very quick to integrate it. We ran a pilot with one of our customers called Stellantis in Europe.
The results of the pilot was incredible. The user satisfaction went through the roof. Usage went up multiple folds. And yesterday, we announced we were the first voice AI vendor to partner with a car company to go live in production-- this is not a pilot-- in 18 countries.
JULIE HYMAN: Like all of the things that we're discussing with AI, especially with images, for example, there's also this sort of uneasiness, right, that you're interacting with something that is not real or not another person, and that there are certain pitfalls of that. How do you think about that issue?
KEYVAN MOHAJER: Yeah. So a lot of issues, like the hallucination is an issue, and copyright, and ethics and all that. And we take those very seriously. The things that we focus on, hallucination is the most important one. And we've worked a lot to reduce that, almost to negligible-- and we're very proud of that.
I think the bigger risk in AI is not embracing it fast enough, right? So there are companies that are kind of disoriented, and they're waiting and seeing-- or they just want to check a box, let's just make an announcement. I think this is a moment to create a lot of value. And the fast-movers are going to be the winners.