What Does It Take to Start a Successful AI-First Company?
Aravind Srinivas, CEO and co-founder of the AI-powered search engine Perplexity, shares his journey and insights into building an AI-first company.
The Journey to Building an AI-First Company
For Srinivas, the journey began with a movie – “Pirates of Silicon Valley.” In a podcast interview on The Next Wave, he recalls, “I never intended to start any company in the beginning, but then there was this movie I watched. It deeply impacted me.”
Inspired by the portrayal of tech giants like Steve Jobs and Bill Gates, Srinivas aspired to be part of Silicon Valley. Seeking a PhD sponsorship, he connected with a professor who encouraged him to delve into deep learning and research, eventually catching the eye of OpenAI co-founder John Schulman.
From Researcher to Entrepreneur
Srinivas’s entrepreneurial ambitions took shape during his internship at DeepMind. Realizing his unique skill set, he said, “I knew I would not be successful starting the next Instagram or TikTok… My skill set is more thinking about some problem deeply and trying to see what we can do with some research.”
Initial Funding and Pivot
Unlike many startups that begin with a clear idea and seek funding, Srinivas and his co-founders received early backing based on their expertise. “Two investors believed in us and gave us $1 or $2 million to play around and tinker,” he explained.
Initially focusing on creating internal search engines for companies, they pivoted after realizing the limited interest. They shifted their attention to larger databases, eventually gaining traction with tech leaders like Google’s Jeff Dean.
The Birth of Perplexity
By expanding their scope to search the entire web, Srinivas and his team developed Perplexity. “Why don’t we just search over the whole web, make the LLM, just look at the links, take the relevant parts, and then let the LLM do all the reasoning?” he described the pivotal moment.
Revolutionizing Search with AI
Understanding the evolving landscape of search, Srinivas highlighted the advantages of AI, “Large language models can help you build new search experiences that were never possible before.”
He acknowledged that while Perplexity is often compared to Google, their platform serves different needs. “They’re going to come to you for asking ‘San Pellegrino or La Croix, what should I get?’ It’s going to register in their mind why you’re different and better.”
Building the Ultimate Knowledge App
Srinivas aims to make Perplexity the go-to app for knowledge seekers. “If people go to Perplexity, they should just feel smarter every day,” he envisioned.
Challenges with Content and AI
Discussing the relationship between content creators and AI companies, Srinivas warned against corporate greed. He suggested a model similar to Spotify, where value is shared among creators, the platform, and users.
“Your bot is only as useful because it’s surfacing good content from the web,” he emphasized the symbiotic relationship between AI and content creators.
The Future of Perplexity
Despite recent announcements about introducing ads, which disappointed some users, Srinivas reassured that this was always part of their plan. Only time will tell if Perplexity can balance user experience with profitability.
