Perplexity CEO on 'Internal Knowledge' AI search, AI race and competition

Financial Freedom
10 followers
4 Video Views·Oct 18, 2024
Aravind Srinivas, Perplexity AI CEO, joins 'Closing Bell' to discuss AI search function, the AI race and competing in search and advertising.
Much of the talk around generative AI is centered around OpenAI, but there is another fast-growing
player in that space who wants a much bigger piece of that pie, Perplexity AI CEO Aravind
Srinivas joins us now, his company ramping up its efforts now to take on Google and others
in search and advertising. Aravind, welcome. It's good to see you and welcome you to our program.
Thank you so much for having me here, Scott. Great to be here.
So you have some new features that you're introducing today which you think will give
users better and more targeted search results, correct? And also deliver them faster. Can you
tell me about that? Yeah, 100%. So today we launched our internal knowledge search. So we
want Perplexity to be the central knowledge base and research platform for everybody. And when
people do their research at work, they're not only requiring research for the data that is already
out there on the internet, on the web, a lot of research is about like due diligence documents
or stuff that's particularly relevant to the company that you acquired in a proprietary way
and storing within your internal knowledge base. And today there exists no one tool where you can
do consolidated centralized research on both these data sources, the web and your internal
company proprietary data together and use both of these for answering your questions or like doing
your diligence or like deciding which companies to invest in all those sort of decisions that you
do on a day-to-day basis. There doesn't exist one single tool to do all of it together and
so far that the technology stack for external search and internal search have been pretty
fragmented. They're very different, they're completely different tools and you have to like
keep moving around data everywhere. It's very inconvenient to do all this but with the advent
of AI, like generative AI, large language models, multi-step reasoning, tool use, it's almost like
every single source of data is just like another API call or a data API call or integration and
these models can decide like what to use when for depending on your query you ask and then we're
able to bring it all together in one platform through perplexity and that's our announcement
today. Do you know what I think is really interesting for our viewers and those who work in
financial services industry is that you're specifically targeting financial services as
your largest vertical here. Tell me about what benefit that's going to be to those who are
either in the financial services industry who are watching now and who are going to use these tools
or for the viewers who are investing in the market and who are going to be better off by
virtue of the tools that will be used by financial services professionals. 100 percent. So we before
today's launch we also broadened the ability to query all the stocks on the market, get news
relevant to any particular stock or like search over related stocks and like look into the 13 apps
filed by big institutional funds. All that data is already being brought onto perplexity for just
generic financial searches and now you have the ability to like take all your internal data and
search over the web. So if you are deciding if you want to like invest in a particular company
and you have all your diligence documents prepared by your analyst and you want to take that and also
query the web based on what people on the web were talking about the company and come up with
an aggregated summary and highlights and like what you need to do and like depending on your portfolio
of your fund should you go and invest in this particular stock. Does it have the right risk
reward ratio. You can ask these sort of questions and these decisions are worth like several millions
or even billions of dollars. So it's completely worth like using a tool to like get you more
clarity. Right. And we're also going to bring in third party data integrations crunch base
pitch book facts that these are all the kind of like data sources that we're looking into
and trying to get that also onto perplexity so that you also get a sense of what the private
markets are investing in not just the public markets. Interesting. I mean obviously we're
very familiar with with all of those. I mentioned at the outset I sort of framed this about your
company in part. You think that Google is vulnerable in search and advertising. Do you
do you think that the duopoly that's existed between Alphabet and Meta truly is is finally
cracking and that you can have a place in that conversation that's meaningful.
100 percent. So it's all about what the consumers are doing right. Consumers earlier used to go to
search engines or social platforms and that's where advertisements were served to them. Now
consumers are increasingly shifting to more native apps. So they're trying to. There's a new consumer
behavior of asking questions and getting answers not just typing in keywords into a search engine
and like getting referred to another link. So that consumer behavior is growing while the existing
consumer behavior of like just typing in a few keywords that sort of saturated. So whenever you
have a new consumer behavior that's on the increasing trend it's always lucrative for
advertisers to figure out how to like be part of that and try to figure out a way to get users
attention in a very different way than what was possible before. And every single query on
perplexity on average has 10 to 11 words. Every query on Google has like two to three words. So
you have the higher intent for every query and the user is like able to ask much more targeted
questions which makes it even more easy for the advertiser to get the right user for their product.
