Aravind Srinivas
CEO & Co-Founder, Perplexity AI
Building an AI-native search engine to challenge Google — the most serious threat to search dominance in over two decades.
Credentials
CEO and co-founder of Perplexity AI. PhD from UC Berkeley in computer science (worked on large language models). Previously research intern at Google (worked with Jeff Dean), OpenAI, and DeepMind. Published researcher in NLP and reinforcement learning.
Why They Matter
Srinivas is leading the most credible challenge to Google Search since Bing. Perplexity doesn't just return links — it reads the web, synthesises answers, and cites sources. For business leaders, Perplexity represents a fundamental shift in how customers find information: instead of ranking your website in a list of blue links, AI will either cite you as a source or you'll be invisible. Understanding this shift is urgent.
Positions
AI Timeline View
AI is already transforming knowledge work. We don't need AGI to revolutionise how people find and use information — current models are good enough if you build the right product around them.
Safety Stance
Key Beliefs
Search is fundamentally broken — ranking web pages is an outdated paradigm. The future is AI that reads the internet for you and gives you answers with citations.
Perplexity founding thesis and public interviews
Citations and source transparency are essential for AI-powered information systems — the user should always be able to verify where information comes from.
Perplexity product design philosophy
Startups can compete with trillion-dollar companies by being faster and more focused. Google's size is its weakness in the AI transition.
Various podcast interviews and public talks
The best AI products are built by small teams that deeply understand both the models and the user experience.
Perplexity's lean engineering approach
Controversial Take
Srinivas and Perplexity have been accused of scraping and summarising content from publishers without adequate attribution or compensation. Major outlets including Forbes and Condé Nast have complained that Perplexity reproduces their journalism while diverting traffic away from their sites. Srinivas argues that Perplexity sends more engaged traffic to publishers than Google does, but the tension between AI-powered search and content creators' livelihoods is far from resolved.
Track Record
How well have Aravind Srinivas's predictions held up?
AI-powered answer engines would become a viable alternative to traditional search within 2 years
Made: 2022
Perplexity reached 100M+ monthly queries by 2024 and is valued at over $9 billion. Google responded by launching AI Overviews.
Google would be forced to fundamentally change its search product in response to AI-native competitors
Made: 2023
Google launched AI Overviews in 2024, fundamentally changing the search results page for the first time in decades — a direct response to Perplexity and ChatGPT.
Key Quotes
“Google is an ads company that happens to have a search engine. We're a search company. That's the difference.”
“The best search experience is one where you never have to click on a link. You just get the answer.”
“I used to think I'd be a researcher forever. Then I realised the best research is building products that millions of people use.”
“Every answer should come with a receipt. If an AI tells you something, you should be able to see exactly where it got that from.”
“We're not trying to kill Google. We're trying to build what search should have been all along.”
Publications
XTX: Cross-Transformer Explainer (Google Research)
2020
Connections
Agrees With
Disagrees With
Sundar Pichai
on Whether the traditional search model (ranking links, showing ads) is still the best way to deliver information — Srinivas says it's fundamentally broken
Gary Marcus
on Whether current LLMs are reliable enough for factual information retrieval — Srinivas believes grounding with citations solves hallucination
Last updated: 2026-04-12
←Back to AI Minds Directory