"We're giving everyone the power that billionaires have."

In an extensive interview published October 18, 2025, Perplexity AI CEO Aravind Srinivas outlined his vision for making elite-level assistance accessible to everyone through artificial intelligence, declaring: “We want to give that power that billionaires have access to to everybody.”

The founder and CEO of the $20 billion AI search startup sat down with tech journalist Tiffany Janzen to discuss Perplexity’s evolution from revolutionizing search to building Comet, a personal AI assistant he describes as humanity’s “second brain.”

The Billionaire Problem: Access to Assistance

Srinivas articulated a stark divide in how the wealthy and ordinary people manage daily tasks. “This power is already available to billionaires or people who are very well off elite society because they have assistants who do this for them,” he explained. “They have employees who do this for them. They have a team of people working with them doing this for them. But the normal person doesn’t have access to all this.”

The CEO recounted a telling encounter at a San Francisco barbershop: “Another old man walked in and he said, ‘Hey, I’m frustrated with my morning. I spend like 3 hours just looking for a new washing machine to buy because my existing one doesn’t work.’ Imagine these are the kind of things that—3 hours because it’s a big volume purchase for most people.”

This anecdote crystallized Perplexity’s mission. “The normal person,” Srinivas noted, still has to “schedule their own hospital appointments. They have to find the best doctors that are covered in their insurance. They have to find local experiences when they plan a trip. They have to find a good flight deal or hotel deal. Very basic things that you’re doing on a daily basis. You don’t even think about it and you spend hours and hours on it.”

Comet: The Second Brain Architecture

When asked to define Comet, Srinivas positioned it beyond conventional productivity tools: “Second brain doesn’t mean a dump of data. I’m not talking about a memory store. Brain is not just meant to be a dump store for memory. A brain that actually can think with you, another brain that can think with you, and you can delegate what your first brain, the core brain finds boring and mundane and tiring, delegated to the second brain.”

He outlined specific use cases: “Stuff like booking reservations, moving around meetings, scheduling a common time for four people to meet, sending reminders to people, getting prepared for the day—let’s say you’re going to interview someone, prepare, looking, reading all their past episodes and podcasts and trying to ask questions that are new and different.”

The philosophy centers on liberating human creativity. “Our first brain can be truly at our natural best selves if we can just be curious and explore and interact and meet people and strategize while the second brain takes care of all the boring mundane workflows,” Srinivas said.

Real-World Applications Exceed Expectations

Since launch, Comet users have found surprising applications. Srinivas shared several examples: “One user was frustrated talking to the customer support of FedEx. He just had Comet talk on his behalf to the customer support which could have been a bot too. Comet got access to the tracking ID and all that and Comet is filing the complaint on his behalf and then going back and forth with the customer support agent on the other end of FedEx.”

Another user “figured out a way to run a marketing campaign on Facebook ads platform for a new product that they were going to put out for a small business that they were running,” while others employ Comet to “unsubscribe from spam emails” by having it analyze sender information and automatically opt out of unwanted lists.

Srinivas envisions more autonomous capabilities: “People are using it to keep up to date on the stock price. You can just say anytime—I don’t want to keep looking at the S&P all the time but anytime there’s a crazy movement just let me know.”

The Vision: An Operating System for Life

The CEO’s ambitions extend to background automation. “The next set of things we’re working on is Comet is running on the background on the server even as you sleep and you don’t even have to have your MacBook or Windows computer open with the browser open. It should just be running on the background,” he explained.

He cited a future scenario: “You might have booked a flight and while you’re sleeping if the flight price actually reduces and Comet is running on the background and it’s going and rebooking the flight for you, canceling your existing booking and saving you like a thousand bucks, then it’s worth it.”

Srinivas framed this as creating “the OS, the operating system” for human life. “The computer science definition of an operating system is where processes are run and memory is being managed, and your life is that except there is no OS for your life,” he said. “You’re still the one that’s orchestrating your life. I don’t think about it as you giving up your agency, but more that delegate the boring aspects of your life that make your life kind of annoying and stressful to something like Comet.”

Academic Roots and Equalizing Knowledge

Drawing on his Berkeley PhD background, Srinivas connected Perplexity’s mission to democratizing academic inquiry. “I’ve always thought like why should academics be the only ones who are allowed to ask questions and think about what is potentially possible and engage in deep scientific research on that. All the smartest people have always been curious. They don’t have to be academics but most people don’t have the platform or even the tools to engage in asking questions.”

He criticized conventional academic structures: “At least during my times at Berkeley, it was not exactly like that. People were definitely after publishing a certain number of papers, building a profile for themselves so that you go get a job in another university as a professor or postdoc or you get hired at Google or OpenAI or whatever.”

The Capitalist Case for Equality

Addressing the apparent tension between profit motives and egalitarian goals, Srinivas was direct: “Of course it’s done in a capitalistic way. We are a profit-minded company but it’s one of the most equalizing things you can do in terms of making something very special and exclusive and elite and giving it in the hands of normal people so that everyone can be the best version of themselves.”

He emphasized time liberation over mere productivity: “We think about it less as productivity and more as making time spent more pleasant and enjoyable. Just do things you enjoy and you’re naturally going to be more creative in that aspect then you’re happier and can do things you enjoy and that’s where real innovation typically comes out.”

Hardware Implications and Future Roadmap

On technical infrastructure, Srinivas discussed both data center efficiency and the potential for on-device AI. “At least in a year or two from now, there is a possibility that a GPT-4 or Gemini or the best models today, a model to that class running on your MacBook in a year or two from now is possible,” he noted, highlighting privacy implications: “If the intelligence, the model can run locally on your computer, we can guarantee full privacy. All your data that gets used by the agent, which is essentially a reasoning model, does not have to go to any server.”

For immediate priorities, Srinivas outlined three focus areas by year-end 2025: mobile optimization for iOS and Android, enhanced voice interaction capabilities, and background automation that functions “even as you sleep.”

Decision-Making Philosophy

Asked about navigating AI’s rapid evolution, Srinivas invoked Amazon founder Jeff Bezos’s framework: “When the world is changing fast you have to ask the inverse question which is what is not guaranteed to change. In his case he asked like in 10 years from now would people want slower package delivery or people want worse customer support or people want less selection of choices? No, they only want more. So work on those problems.”

Applying this logic, he concluded: “People are always going to want faster answers. People are always going to want more accurate answers. People are always going to want the AIs to do things for them, not just answer stuff, but actually go do stuff for them. So you got to work on these problems regardless of what happens in terms of whether the models are getting cheaper or more expensive.”

On leadership lessons from successful entrepreneurs, Srinivas emphasized: “Everyone has one common aspect: resilience. That’s the most important characteristic that I try to take from that because things will not always go well and I’ve had so many days in which just waking up felt so miserable. But then the whole company’s working here, the investors have put a lot of faith in me, the employees even though they’re doing their job that’s scoped out for them, fundamentally they have a lot of stock and they’re relying on me.”

The interview concluded with Srinivas’s ultimate aspiration: “The bar is you got to be able to feel the utility of the product to the extent that your work and life should run on it. That’s why I call it the OS, the operating system, because your life is that except there is no OS for your life.”