Loading summary
Transcript110 lines
- [00:00]
Palantir Narrator
You're being sold an AI future where you're obsolete or irrelevant. That vision is wrong. At Palantir, they're building AI that helps workers and unlocks their full potential. American workers are our nation's greatest strength. AI shouldn't eliminate them, it should elevate them. Palantir is here to tell their stories. From factories to hospitals, AI is freeing people from drudgery, letting them do what humans do. Create, solve, build. Palantir Making Americans irreplaceable Being a small.
- [00:36]
Tracy Alloway
Business owner isn't just a career, it's a calling. Chase for Business knows how much heart and effort go into building something of your own. Manage all your business finances, from banking to payments to credit cards, all in one place with their digital tools. Plus access online resources designed to help your business thrive. Learn more@chase.com business chase for business Make More of what's yours the Chase Mobile app is available for select mobile devices. Message and data rates may apply. JP Morgan Chase Bank NA Member FDIC Copyright 2025 JPMorgan Chase & Co.
- [01:14]
Dmitry Shevalenko
Did my card go through?
- [01:16]
Joe Weisenthal
Oh no.
- [01:16]
Dmitry Shevalenko
Your small business depends on its Internet, so switch to Verizon Business and you could get LTE business Internet starting at $39 a month when paired with Select Business Mobile Plans. That's unlimited data for unlimited business. There we go. Get the Internet you need at the price you want. Verizon Business Starting price for lte Business Internet 25 Mbps Unlimited Data Plan with select Verizon Business Smartphone plan. Savings terms apply.
- [01:45]
Tracy Alloway
Bloomberg Audio Studios Podcasts Radio News.
- [02:00]
Joe Weisenthal
Hello and welcome to another episode of the Odd Lots Podcast. I'm Joe Eisenthal.
- [02:05]
Tracy Alloway
And I'm Tracy Alloway.
- [02:06]
Joe Weisenthal
Today you are going to be listening or watching a special episode of our podcast recorded live at a conference which you've probably never seen before.
- [02:16]
Tracy Alloway
Yeah, this was at the Lazard Foursquare conference. And we interviewed Dmitry Shevalenko. He is the chief Business Officer of Perplexity, which is something you might have used before. We certainly use it.
- [02:26]
Joe Weisenthal
Yep, one of the leading AI models. And so this is a conference related to media and deal making. And so we talked about what the impact of AI is going to be on news media and brands, the Internet as we know it.
- [02:39]
Tracy Alloway
We also took questions from the audience so you will hear us referring to questions that we are receiving on the app. We hope you enjoy. Take a listen. All right, so here's my first question. Should we even be talking to you? What I mean by that is I could just go on Perplexity and ask all these questions and it'll give me a pretty decent answer. What's the point?
- [03:01]
Dmitry Shevalenko
First of all, thank you for having me. It's great to be here. I love that tee up and you guys were scheming backstage of what's going to be the hard opening question. But I love it because the thing that humans will always be uniquely exceptional at is asking questions. So perplexity may have the answer, but perplexity does not have an innate desire to be curious. And that's what you guys have. And that's why journalism is so important and that's why having these conversations important. It's about the questions, not necessarily the answers.
- [03:41]
Joe Weisenthal
Kind of related to this, I guess. But here's the thing that I always think about, which is I enjoy using perplexity, et cetera. Let's say you have great relationships with all the news publishers, et cetera, and they're really cool with it all and all those deals get solved, et cetera. Why do we need the Internet as we know it that has things called websites built for humans now that we're in this age where we can go to a chatbot or something like it and. And the answer is right there. Do we need to have what are traditionally called websites anymore? Why would a publisher continue and invest in design and all of these things when I could just get the answer from Perplexity or chatgpt or whatever else?
- [04:24]
Dmitry Shevalenko
So again, it goes back to. I think the reason we're drawn to media properties is because of the questions the journalists and editors are asking. Right? It's not necessarily. It's the framing and it is the what is actually worth asking about, what's worth reading about. Perplexity doesn't have answers to that. You have to come to perplexity already armed with a question. And oftentimes people read news and there's something they didn't understand fully and they come to Perplexity to get that deeper understanding. But the spark that comes from editorial judgment, that comes from the intuition of a world class reporter. And that's why we're excited through programs like Comet plus to find new business models for that.
- [05:10]
Tracy Alloway
How do you actually judge editorial judgment or quality of news sources? Because I can imagine a future where a lot of our news consumption is through something like a perplexity model, which means that you're basically the arbiter of what information is reliable. So how do you make those decisions?
- [05:28]
Dmitry Shevalenko
It's not me at the wheel. Ultimately, this is where we focus a big part of our technology. It's building accurate AI. The thing that we believe that's going to be most scarce in the future is not intelligence, it's trust. I think the bedrock of trust is transparency. That's why Perplexity pioneered showing you exactly which source the information is coming from and empowering users to apply their own judgment of whether that is a good source or not so good source. But we deploy many different algorithmic strategies. Oftentimes, if many publications are saying the same thing, that's a strong signal that they're onto something. And the outliers you're kind of more sensitive to, but it's not a simple problem. And this is why we raise a lot of capital and hire the world's best technologists to build accurate AI.
- [06:27]
Joe Weisenthal
So I think anyone could sort of intuitively understand why publishers might be anxious about Perplexity as being a destination. Evidently, it's not just news. Publishers, literally. Since we got on stage three minutes ago, news broke that Amazon demands perplexity to stop AI agent from making purchases. Just broke 3 minutes ago from Bloomberg.
- [06:53]
Dmitry Shevalenko
What is the good job, Bloomberg?
- [06:54]
Joe Weisenthal
It's not just news that's anxious about the idea of Perplexity being the destination. Clearly it's all kinds of businesses.
- [07:02]
Dmitry Shevalenko
So I think, and I haven't read the article yet, but I'm going to read it out, I think I have an understanding of the issues. Ultimately, the issue with Amazon is tied to Perplexity's web browser called Comet. And Comet is a web browser that has an AI assistant built into it. And what we think is important is that consumers and end users have access to powerful AI when they are accessing publications, accessing websites, accessing the Internet. The first 30 years of the Internet, we had lopsided AI, where big tech, Google, Netflix, Facebook, they used AI to control what we see, to get us to click on certain things and not click on other things. And what is changing now is, is Perplexity is focusing on giving powerful AI to the end user. And we think this is the right thing to do. We think it's a principled viewpoint that when you're on Amazon, it's not Amazon's business whether you're using AI to help you evaluate. Is this the best thing to buy? Can I find it cheaper somewhere else? Do I actually need to buy the other thing they're telling me to buy with it? The world's wealthiest people, they have personal shoppers that help them do this. And we're democratizing that sort of intelligence, that sort of assistant access to everyone.
- [08:26]
Tracy Alloway
Just in terms of competition. Is the argument basically that technically, you know, a Google or an Amazon could create similar products to what you guys are doing, but their own business model basically prohibits them from doing what you're doing.
- [08:42]
Dmitry Shevalenko
I should have you join our next investor pitch. It's exactly that. Oh, dear, it is. There is a big opening for an independent, neutral AI player that is aligned with users. Google is still. They're generating 100 billion in revenue per quarter, which is amazing. But a lot of that is coming from getting their customers advertisers. Amazon, they're trying to get you to buy as many things as possible. Our success metric is do users, are they willing to pay for a perplexity subscription? And so we're purely aligned with our users.
- [09:25]
Joe Weisenthal
Speaking of competition, like, what is the differentiator? Okay, maybe Amazon has some obvious angles and some of the others OpenAI seems sort of independent in the same bucket. It's really huge. It's one of the biggest websites in the world. It doesn't strike me. Many of the things that they offer replicate seem sort of similar to perplexity. You must get this. I know you have private stock answer this, but what is the differentiator between you and an OpenAI?
- [09:55]
Dmitry Shevalenko
So I'm fortunate to get to speak with a lot of curious people that have to answer a lot of hard questions every day. And they tell me they use perplexity. And it comes back to we're focused on accuracy above all else. ChatGPT is an amazing product, but they built a product that is optimized for engagement. It's trying to be your cheerleader, your coach. It's trying to get you to keep going. And they've also built a product that is designed not to use the best LLM in service of your question, but it's designed to use their LLMs. And so part of being neutral, part of being independent, trustworthy, accurate, is using the best possible technology to answer the question the best possible way. And so being an aggregator of all the best models, having our own independent search infrastructure that isn't The Bing Search API, which is what ChatGPT uses, those are all the building blocks of a more useful, accurate AI.
- [11:00]
Tracy Alloway
What's your actual revenue source at this point? Because I know you're big on subscriptions, you experimented with ads, I think you might have paused that at this point. But it does seem if you're trying to be different to Google, you probably don't want to be as ad driven as they are.
- [11:15]
Dmitry Shevalenko
Yeah. So we're seeing tremendous traction with both consumer subscriptions. And then companies buying the enterprise version of Perplexity for their employees. Ultimately it's in that enterprise context where having accurate, useful knowledge is incredibly valuable. And so we've been fortunate with an enterprise go to market team of just five people. Seen that business more than 8x this year, seeing many multiples increase in our consumer subscriptions. We're excited about building products that are user aligned, that also create marketing value for brands and advertisers. But this isn't going to be rebuilding AdWords and click based systems. We think it's important to create real value so that when your personal AI agent powered by Perplexity is evaluating a potential commercial opportunity, something you might buy, imagine you know alongside the answer, you got an offer for an exclusive discount. That would be something that your agent would say, hey, this is. Even though this, someone's paying for this to show up here, this is actually a good deal. If it's just an ad that says click me, click me, click me, we don't think that's going to be sustainable. And that's not the type of marketing products we're looking to build.
- [12:52]
Palantir Narrator
Silicon Valley is selling you a future where you're obsolete or worse, identical. At Palantir, they're witnessing something different and revolutionary. From re industrializing the nation's defense base to shipyard workers building faster and frontline workers, boosting productivity. AI is transforming work across the nation. AI is not replacing American workers or flattening them into conformity. It's unleashing what makes each one their judgment, their craft, their creativity. When American workers become more powerfully themselves, they own the future. Palantir making Americans irreplaceable.
- [13:33]
Dmitry Shevalenko
There are two kinds of people in the world.
- [13:35]
Tracy Alloway
People who think about climate change and people who are doing something about it. On the Zero podcast, we talk to both kinds of people, people you've heard of, like Bill Gates.
- [13:44]
Dmitry Shevalenko
I'm looking at what the world has to do to get to Zero.
- [13:48]
Joe Weisenthal
Not using climate as a moral crusade.
- [13:51]
Tracy Alloway
And the creative minds you haven't heard of yet. It is serious stuff, but never doom and gloom. I am Akshat Rati. Listen to Zero every Thursday from Bloomberg podcasts, on Apple, Spotify or anywhere else you get your podcasts.
- [14:04]
Joe Weisenthal
Can we talk a little bit about the fundraising environment for AI? I mean, obviously we know how crazy, you know, intense things are in the.
- [14:10]
Tracy Alloway
World are people throwing cash at you as you walk down the street.
- [14:13]
Joe Weisenthal
But also it feels like every day. It used to be when I paid attention to startups, it felt like there were discrete rounds There was a C, then an A, and then a B, and then maybe a C and then an ipo when I read it feels like every day there is just are the rounds blending into each other so that it's just this sort of ongoing stream of oh, here's a new raise, et cetera. Here we got some new chips and then we are giving them some equity, et cetera. Has the line between rounds, is it blurring?
- [14:41]
Dmitry Shevalenko
Yeah. So before Perplexity, I mean, I spent most of my career at consumer Internet companies by this stint as a founder of a robotics startup. And AI is definitely a different fundraising environment than that was, which was very challenging. I think at its core, a lot of investors believe, and we agree with this belief, that what's happening with AI now is as profound as the original introduction of the Internet. And so there are going to be a wave of new generational companies that are being built right now. And a lot of the value is actually compounding while these companies are privately held. And so that is, there's a lot of conviction by private investors in that. I think a second piece is that investors themselves are power users of products like Perplexity. They are people that have a lot of questions, do a lot of research. And so they have not just an abstract understanding of a balance sheet, they have their own intuition, which is very different than like B2B SaaS where it's like you have to go do 30 customer interviews and you're kind of getting all the second order information with Perplexity. Use it yourself. You understand the value, the time that you just saved, the creative expansion you just had. And that gives investors conviction. And so that is fueling this environment. Is it at times too frothy? Sure. Are there going to be some investors that make the wrong bets? Absolutely. But the core human need of getting good answers to good questions, we think that is boundless. And so we're confident in the path ahead.
- [16:33]
Tracy Alloway
Well, speaking of froth, is there anyone you turn away in terms of being an investor?
- [16:38]
Dmitry Shevalenko
Yeah. So there's this phenomenon that tends to happen in bubbly periods of multi layered SPVs. So SPVs are special purpose vehicles. And so you basically have many layers of outsourced fundraising where you have principal agent misalignment, where the fundraiser is not making money off the investment, they're just making money off the management fee. And so we often get approach from players in that space and whenever we can detect that's what it is, we avoid taking on funds like that because we want investors who are getting direct Access to perplexity and who are in it for the long haul, who aren't just trying to make a quick buck on a management fee flip.
- [17:26]
Joe Weisenthal
I want to talk about this more. I'm really fascinated. We did an episode with Vlad 10 of the Robinhood CEO and he was talking about tokenization of private assets. And I guess this sort of actually relates to the question about discrete rounds. But does it feel like in theory if some sort of slice of equity perplexing, maybe people sell forwards or these vehicles that are linked to it. Does it feel like the line between what it even means to be a private and a public company is inevitably going to disappear? Especially add in the fact that we seem to be on this road where public companies will not be expected or required to have to report information as timely as they used to.
- [18:06]
Dmitry Shevalenko
I think it's important when you're a privately held company to control the sale of your equity. I think that the challenge why we haven't. We've been approached by a lot of these tokenization schemes and again, if you have folks on your cap table that are not aligned for the long term, they may be looking for quick flips and then that creates a weird signal in the market and that may harm a primary raise. Right. And so the noise that comes with that, it's not like we're short on other investments avenues. So it's just an unnecessary kind of noise risk.
- [18:47]
Tracy Alloway
So speaking of throwing money around, the $35 billion bid for Chrome, what was that? Was that a troll?
- [19:01]
Dmitry Shevalenko
When we made that bid, the DOJ had still not decided whether Chrome would be divested from Google. The web browser, as evidenced by Amazon's behavior, is clearly incredibly valuable. Surface area 35 billion was the starting bid. At this point I can safely say I think Chrome is worth more than that. But what I can confidently say is we had more than 35 billion of investor interest in making that bid. So it was not a constraint on the funding side. Obviously Google does not want to sell Chrome unless they'll be forced to, and it doesn't look like the government will be forcing them.
- [19:45]
Tracy Alloway
So you're serious about that one? It's not just a sort of antitrust flag waving exercise?
- [19:50]
Dmitry Shevalenko
No, we had very large sovereign wealth funds that were kind of lined up and ready to go.
- [19:57]
Joe Weisenthal
Here's a great question from someone anonymous via the app. Can you talk about unit margins? You often hear this. It's like, oh, these AI companies, they're not even on a unit basis making money, given the compute costs, etc. Can you talk about gross margins and right now whether like you know, a given query is profitable given the cost of delivering.
- [20:21]
Dmitry Shevalenko
So the a paying pro subscriber to perplexity is profitable for us. Okay, so the unit margins are strong. The free users right now because we do not monetize them. We lose money on those AI companies, I think something that is underappreciated. They generally actually when they do the accounting of free users, they attribute that to research and development because you're still using those prompts to improve and train.
- [20:55]
Joe Weisenthal
Your system, the model.
- [20:57]
Dmitry Shevalenko
So it does allow for your unit economics to potentially appear more pristine than they are.
- [21:03]
Joe Weisenthal
Are all pro users profitable? Because this also strikes me as something unique with AI, which is that, you know, when back when Twitter was in its infancy, power users were good. It feels like in the case with AI, sometimes you get these power users where you're losing money on them. Are all pro users profitable or do some use it so much that it's more than the subscription fee?
- [21:22]
Dmitry Shevalenko
I mean it's not a material amount. And we've also introduced a max tier, so we have a $20 a month pro subscription. And for the folks that want the most expensive reasoning models and higher limits on them, there's a $200 a month product called max and that also gets you access to our email assistant. And so I think you'll and other companies have also introduced kind of higher tiers of subscription. So this is not a problem that's unique necessary to AI and you kind of introduce tiers of service to support it.
- [21:56]
Tracy Alloway
At scale, what does your expenditure actually look like at this point? What are you spending the most money on? We read about these insane salaries for engineers going to Claude or OpenAI or whatever. Is it talent retention or is it compute power? Is it the content deals with media organizations? What are you spending on?
- [22:17]
Dmitry Shevalenko
The biggest expense is inference. It is the cost of compute when you're asking a question and running through an LLM that is a material expenditure. Perplexity doesn't do pre training. So the sorts of AI PhDs that other companies are paying reported hundreds of millions of for. We're not in the market for that. We have incredible AI application engineers and that's what we recruit for. Ultimately, engineers that are joining Perplexity are joining because they believe we're just getting started, right? And they're coming for an equity appreciation story that's already bearing fruit. But we're just at the beginning of that journey.
- [23:06]
Joe Weisenthal
Maybe this is sort of a way of restating Tracy's question. But it all feels like in any AI conversation, it's like, what is the bottleneck right now, right? Is it talent, is it capital, is it computer, is it power, et cetera? Like, if you could have unlimited amounts of any one of those things, what would you want more of right now?
- [23:27]
Dmitry Shevalenko
So I think it's important to bifurcate. There's the companies that do the training of foundation models and those are the ones that are really pushing for the massive infrastructure buildouts. We're not constrained on compute for inference at the current usage levels, but as behavior shifts, I mean, by a show of hands in the room, how many people have stopped primarily using Google when they have an informational question? So about 80% of the room raised their hands. And we see the early adopter behavior in audiences like this, folks that work in financial services, people that again, for whom quickly getting high quality information matters. As this behavior becomes more mainstream, we obviously will need more compute for inference. But the folks that are talking about trillions of dollars in investment, it's really a different universe than ours. It's unclear to me why some of those investments are needed. Sometimes it feels like a strategy of, you know, kind of what, what the US was trying to do with the Soviet Union of just getting your competitors to go broke by, you know, being able to raise the most capital. And we'll see how that plays out.
- [24:50]
Tracy Alloway
Since you mentioned financial services, we have a good question from the audience. We've done some media navel gazing. I'm sure we're going to do some more. But for now, financial services navel gazing. What do you think the impact of AI will be on places like Lazard and other Wall street firms?
- [25:08]
Dmitry Shevalenko
I just think this is going to be a incredible flowering of creativity and productivity lift. I think of using a product like Perplexity and other AI tools in the day to day of somebody that works in an investment bank is really twofold. The first is saving you 80% of the time to get to the 80% first draft of a work product. And investment banks, that's kind of putting together pitch decks, doing modeling, just gaining tremendous efficiency. It's important to distinguish, it's not about making that workflow autonomous. You still absolutely need human judgment on what the end output should be, how to approach it. This is not about replacing investment bankers, it's about giving them tremendous leverage. I think the more disruptive impact is now that it got a lot easier to begin a new initiative, to begin a new prospect. How many more new engagements are you going to be able to Spin up. Right. Because the biggest constraint on new business is an employee feeling they have the bandwidth to take it on. And experienced leaders now have much more leverage. And I think that's going to play a profound role at Lazard and part of why I'm excited to be part of the organization.
- [26:45]
Joe Weisenthal
Five times a day, CNN brings you five stories that'll get you up to speed on your day. New episodes drop Monday through Friday at 6:00am, 9:00am noon, 3:00pm and 6:00pm hello from CNN.
- [26:59]
Dmitry Shevalenko
I'm Joe Beck from CNN. I'm Fez Jamil.
- [27:03]
Joe Weisenthal
I'm Christa Bow with the five things.
- [27:04]
Tracy Alloway
You need to know.
- [27:06]
Joe Weisenthal
Follow CNN's Five Things podcast wherever you get your podcasts. I love these audience questions because they're saying things that I would think that reminded me and that I would have been rude to say. You didn't answer the first question about are you going to kill journalism? But someone says you didn't answer the first question how will journalism survive when AI has skinned it? I get, oh yeah, AI isn't interested in asking questions. But it's not the question is why would brands continue to build out these presence for a human reading and scrolling when I can just go to one destination that doesn't necessarily click me anywhere else?
- [27:49]
Dmitry Shevalenko
Well, I mean when I look at the earnings numbers for the Wall Street Journal, the New York Times for Bloomberg Media companies seem to be doing fine.
- [27:58]
Joe Weisenthal
Well, the New York Times is doing well for games and I don't know if you get earnings numbers for games.
- [28:02]
Dmitry Shevalenko
But what I hear like subscription numbers.
- [28:06]
Joe Weisenthal
But the I should say by the way, units doing a lot more than games. I actually did not. It's a very impressive company but by and large it does not look like the actual news business. It's a lifestyle company is doing particularly well.
- [28:17]
Dmitry Shevalenko
Well but I think it's part of the, the value of the bundle. I mean to me this is very analogous to when the Internet was first introduced. You had traditional brick and mortar retailers and the best brick and mortar retailers adapted to the digital environment and the best publishers are adapting and gaining leverage with AI. I mean last night we had the CEO of Axel Springer talk about how he's leaning into AI and and it's actually an incredible opportunity for productivity. And so I don't think this is zero sum. I do think there's going to be some publishers that thrive and others and the media business was already facing challenges certain segments of it well before AI. So I don't see this as being kind of the fundamental tension.
- [29:10]
Tracy Alloway
But if you Fast forward five or ten years from now. What does the front page of the Internet actually look like to you? And what exactly is, I guess, the economic pact between media and platforms such as yourself? What would the ideal pact or agreement actually look like? Basically, give me your vision of the Internet in five years.
- [29:32]
Dmitry Shevalenko
In two seconds, we're introducing something called Comet plus, which we already have. The Washington Post and Conde Nast, Louisiana Times, many great publishers signed on for. And it was really Inspired by Apple News plus, where if you are a Comet plus subscriber, which is $5 a month, and we bundle it into Perplexity Pro, so even people that didn't necessarily want to pay for it, we're taking money out of our cut and feeding it into this program to seed it. As you consume premium content in the Comet browser, we're then sharing that revenue back proportionally with publishers and it's actually accounting for not just page views of premium content, but, but agent actions on that content and citations in the comment browser. And so ultimately, I think subscriptions are a very powerful business model for media. I think there's a lot more to do there and we want to play a leading role in doing that.
- [30:30]
Joe Weisenthal
I'm very impressed by the finance module within Perplexity. There's a whole graveyard of companies that have tried to take down our employer at Bloomberg. Do you see yourself as a potential challenger company Tracy asked about, you know, financial services, et cetera. Do you see yourself at some point down the road as a challenger to the Terminal?
- [30:54]
Dmitry Shevalenko
Yeah, I mean, how many people have access to Bloomberg Terminal? It's like. I mean, you guys, like, it's not. It's not millions, right? Yeah, it's not millions. So for us, the success of Perplexity Finance is. It's similar to what I was saying about with Amazon and people that have personal shoppers like Bloomberg Terminal. Is that like, it's. This is like.
- [31:13]
Joe Weisenthal
But we know it's very lucrative and we know it's a very big company. And so when you think about the lazards of the world and all the other banks, could you at some point see yourself pitching Perplexity or Perplexity Terminal as where all these other ones had failed to do it?
- [31:26]
Dmitry Shevalenko
I think for us, the bigger.
- [31:27]
Joe Weisenthal
You won't be offended.
- [31:29]
Dmitry Shevalenko
The bigger opportunity is not to go after Bloomberg Terminal. It's giving everyone who can't afford Bloomberg Terminal a version of it that is useful for them. And when they're making retail stock investment decisions, when they're understanding the interplay of the markets and the news cycle, that they have access to powerful technology. Right. So it's not, I don't think we can't like Bloomberg Terminal. Like people are there for the chatting. Like there's all kinds of features there that we're not trying to replace.
- [31:59]
Joe Weisenthal
Thank you.
- [32:00]
Dmitry Shevalenko
And it's an amazing product and amazing brand. I think it's for us, this is about democratizing, expanding the scope of who has access to premium financial intelligence.
- [32:11]
Tracy Alloway
Another question from the audience. This one's from Olivier Chastain. What do you think will be the key skill sets for media entertainment executives to manage the impact of AI on content industries?
- [32:26]
Dmitry Shevalenko
I mean, I think my answer to this is the answer. I think the key skill set for any professional in the age of AI is you have to get really good at asking the right questions. And so that means asking your teams why are they working on what they're working on, what's the leverage, what's the unique differentiation? And so it is approaching their business with a profound curiosity and an understanding that there are new disruption and leverage opportunities and, and to kind of lead with that curiosity.
- [32:59]
Tracy Alloway
Everyone always says that. But what does a good prompt actually look like to you? Can you give us a specific example of like a prompt that really impressed you?
- [33:07]
Dmitry Shevalenko
I think to me asking a good question, it's like a level deeper than like the prompt engineering is kind of that. That in some ways will get abstracted very soon. When I say asking good question, it means a question that embodies and embeds all the known knowledge on a given topic. And the question you're asking is then where do you push the frontier? What is the thing that is not yet knowable, that in your question you're pushing your organization on, you know, going.
- [33:38]
Joe Weisenthal
Back to the browser question and about the bid for Chrome, Google itself earlier this year, there were a lot of people like, oh, Google is going to get disrupted by AI. It's classic innovators dilemma. They invented all this tech and they're not making now the stock market views. Google's done phenomenally. Or Alphabet, they've done phenomenally well. And a lot of people have really changed their tune on them. If the browser is such an important part of the real estate here, does that just give them a massive leg up? I mean, I know you're gonna have Your own and OpenAI is gonna have its own, but Chrome is everywhere. That sounds like a very big moat.
- [34:15]
Dmitry Shevalenko
It's a very powerful distribution leverage point. But their greatest strength is also a weakness in that they built it around a AdWords system where they're reliant on people clicking on links. And the behavior of clicking on promoted links will just matter less in the future of the Internet than it has in the past.
- [34:36]
Joe Weisenthal
Speaking of promoted links, this is another question I've been wondering about. Sometimes I see these companies and they advertise on LinkedIn and they offer what I think is sort of would be the equivalent of like the AI equivalent of SEO. Is that all Snake oil is there? Like, whatever. Does that exist? Can you. Are there these games that publishers or any brands can play to do well in AI? Is that sort of a fraud?
- [34:59]
Dmitry Shevalenko
I don't want to call it a fraud. I have not uncovered those companies provide analytics, but they are not actionable. And so I have not found an actionable output there. And in some ways, our purpose is to prevent those insights from being actionable, because if you can game a Perplexity answer, it means it's not accurate and trustworthy. And so if they do find a hack, we have better engineers that will fill that hack.
- [35:32]
Tracy Alloway
I have a slightly philosophical question, but is there any way to inject a tiny bit of randomness or creativity in some of Perplexity's output? And what I mean by that is, I think at this point, we're all aware of algorithmic content moderation and the idea that if you start your Netflix account and you watch one romantic comedy for the rest of your life, Netflix is going to be throwing, like, Matthew McConaughey films at you forever and ever and ever. How do you deal with that aspect of it? The risk that people are funneled into, like, one particular answer stream or one corner of the Internet.
- [36:11]
Dmitry Shevalenko
So memory will be an important feature of products like Perplexity. And the way we avoid that trap is transparency, just like we do with sources where we actually show you all of the things that we've added to your memory. And if you don't want to be pigeonholed as a romantic comedy lover, you just delete that memory. And I mean, I'm kind of. I recently got into health supplements, and now every time I ask about health supplements, my decision to begin pomegranate supplements is getting mentioned. The answer.
- [36:43]
Joe Weisenthal
Oh, I get it.
- [36:45]
Dmitry Shevalenko
Yeah. And I del. Deleted that because I'm like, you know what? I don't want it coming back to the pomegranates, as great as they are. And I just went and deleted that memory. I saw where it was and I'm pomegranate free in my answers.
- [37:05]
Joe Weisenthal
There's a question here about let's See the comparison to the first wave of the Internet. So this person is talking about content creation, AI, agent, studios, pace of change moving very fast. I mean, what's really extraordinary to me about the last several years is the speed with which we seem to get new paradigms right. Blogs are going to be a big deal. That was 20 years ago. They had a lifespan of a few years. Then social media, Twitter, et cetera, their speed. How do you think about things get overturned so fast these days? How do you think about making investment decisions or under such conditions of uncertainty?
- [37:44]
Dmitry Shevalenko
Yeah, this is a funny time to drop his name, but this is where we love the Jeff Bezos line of the hard thing to predict is not what's going to change, but what's going to remain the same. I think from our vantage point, the bedrock is accuracy and trust. And that's actually as intelligence gets much more powerful, it gets more powerful at manipulating us. And so that's why that trust bedrock is paramount. That said, one of my management mantras is that I have absolute conviction that six months from now I'm going to have a top three priority, that today I don't know what it is. And that embedded in that is an understanding that even the model makers themselves, when they release a new model, until it's out, they don't know what emergent capabilities and what the useful applications of that will be. And so kind of having core foundational principles while at the same time preserving agility and adaptability and being the best at applying the new gifts we get for the service of users. That's going to be a key operating skill.
- [38:58]
Tracy Alloway
There's a second part of that audience question that Joe just asked, and I think it's probably the one that's on all of our minds, and it's the bluntest one. But is AI in a bubble?
- [39:10]
Dmitry Shevalenko
There are certainly parts of it that.
- [39:12]
Tracy Alloway
Are not you, obviously, but your competitors, presumably.
- [39:16]
Dmitry Shevalenko
I mean, I think of all of the time and leverage that professionals and knowledge workers are getting, and to me, the ROI on that. Like, I can see the trillions of dollars in economic productivity happening from it, but are there certain companies that have just raised too much capital and is going to create imbalances in the ecosystem? Yes.
- [39:39]
Joe Weisenthal
All right, but we only have 45 seconds. My issue is a lot of things still don't work in AI. I tried to ask ChatGPT yesterday how many full school weeks there are in New York City public schools. It was like, oh, I'd have to count it by hand. And no, it seriously said that I posted a screenshot of it. A lot of these agent things, they don't work very well. A lot of things. I once asked who are the most common guests on Odd Lots over the years, it just gave me a bunch of this was not your site, but I'm just saying a lot of AI agents and so forth, they sort of. About halfway through.
- [40:12]
Dmitry Shevalenko
Yeah. I'm tempted to do a live check of whether Perplexity will give you a good answer on that. But this goes back to what I was saying about emergent capabilities. The fail case and this comes up actually in many enterprise conversations is just because you tried something two months ago, it doesn't mean and it didn't work doesn't mean it's not working today. And so we all need to keep this open mind that these technologies are getting smarter and better at a rapid clip, and those are solvable problems.
- [40:43]
Tracy Alloway
All right, Dimitri, thank you so much for thank you Conversation. Really appreciate it.
- [41:02]
Joe Weisenthal
That was our conversation with Dmitry Shevalenko, chief business Officer at Perplexity.
- [41:07]
Tracy Alloway
Shall we leave it there?
- [41:08]
Joe Weisenthal
Let's leave it there.
- [41:09]
Tracy Alloway
All right. This has been another episode of the Odd Thoughts podcast. I'm Tracy Alloway. You can follow me at Tracy Alloway.
- [41:14]
Joe Weisenthal
And I'm Joe Weisenthal. You can follow me at the Stalwart. Follow our guest Dimitri shevalenko. He's at Dimitri140. Follow our producers Carmen Rodriguez at Carmen, Armand, Dash O' Bennett at Dashbot, and Kale Brooks at Kale Brooks. For more Odd Lots content, go to bloomberg.com oddlots we have a daily newsletter and all of our episodes and you can chat about all these topics 24. 7 in our Discord, Discord GG oddlots.
- [41:38]
Tracy Alloway
And if you enjoy Odd Lots, if you like it when we do these live recordings, please leave us a positive review on your favorite podcast platform. And remember, if you are a Bloomberg subscriber, you can listen to all of our episodes absolutely ad free. All you need to do is find the Bloomberg Channel on Apple Podcasts and follow the instructions there. Thanks for listening.
- [42:01]
Dmitry Shevalenko
Sam.
- [42:29]
Joe Weisenthal
Five times a day, CNN brings you five stories that'll get you up to speed on your day. New episodes drop Monday through Friday at 6:00am, 9:00am noon, 3:00pm and 6:00pm hello from CNN.
- [42:42]
Dmitry Shevalenko
I'm Joe Beck from CNN. I'm Fez Jamil.
- [42:46]
Joe Weisenthal
I'm Christa Bowe with the five things.
- [42:48]
Tracy Alloway
You need to know.
- [42:49]
Joe Weisenthal
Follow CNN's five things podcast wherever you get your podcasts.