Search for a command to run...

Timestamps are as accurate as they can be but may be slightly off. We encourage you to listen to the full context.
This episode of Odd Lots features a live interview with Dmitry Shevelenko, Chief Business Officer of Perplexity AI, recorded at the Lazard Foursquare conference. (02:02) The conversation explores how AI companies like Perplexity are impacting traditional media, the economics of AI startups, and the future of information consumption on the internet.
Dmitry Shevelenko is the Chief Business Officer of Perplexity AI, one of the leading AI-powered search and information platforms. Before joining Perplexity, he spent most of his career at consumer internet companies and had experience as a founder of a robotics startup. He has been instrumental in developing Perplexity's business strategy and partnerships with media organizations.
Joe Wiesenthal is co-host of Bloomberg's Odd Lots podcast and a veteran financial journalist. He focuses on markets, economics, and the intersection of technology with traditional finance.
Tracy Alloway is co-host of Bloomberg's Odd Lots podcast and covers financial markets, economics, and business trends. She brings expertise in analyzing complex financial and technological developments.
Shevelenko emphasized that while AI can provide answers, humans remain uniquely exceptional at asking the right questions. (03:15) He noted that "perplexity may have the answer, but perplexity does not have an innate desire to be curious." This highlights that journalism and human editorial judgment remain valuable because they determine what questions are worth asking and what stories deserve attention. The spark of curiosity and editorial judgment cannot be replicated by AI systems, making human-driven content creation still essential.
According to Shevelenko, "the thing that we believe is that's going to be most scarce in the future is not intelligence, it's trust." (05:40) Perplexity addresses this by showing users exactly which sources information comes from, allowing them to apply their own judgment. This transparency-first approach differentiates them from other AI platforms and builds user confidence. In an era where AI can generate convincing but potentially inaccurate content, transparency becomes a competitive advantage and a foundation for sustainable business models.
When discussing AI's impact on investment banking and professional services, Shevelenko explained that AI tools like Perplexity save professionals "80% of the time to get to the eighty percent first draft of a work product." (26:54) However, he emphasized this isn't about making workflows autonomous - human judgment remains essential for determining what the end output should be and how to approach problems. This creates tremendous leverage for experienced professionals, allowing them to take on more new engagements because the biggest constraint on new business is employees feeling they have the bandwidth to handle additional work.
Shevelenko argued there's a significant opening for independent, neutral AI companies that are aligned with users rather than advertisers or platform owners. (08:24) While Google generates $100 billion per quarter primarily from advertisers and Amazon tries to maximize purchases, Perplexity's success metric is whether users are willing to pay for subscriptions. This alignment with user interests, rather than advertiser interests, allows them to build products that truly serve end users - like helping people evaluate purchases on Amazon or find better deals elsewhere, even when it conflicts with platform owners' business models.
When asked about profitability, Shevelenko confirmed that "a paying pro subscriber to Perplexity is profitable for us" while acknowledging they lose money on free users. (21:49) The company has introduced tiered pricing with a $20/month Pro subscription and a $200/month Max tier for power users requiring expensive reasoning models. This approach to unit economics is crucial for AI companies to achieve sustainability, as compute costs for inference remain their biggest expense. The key insight is that successful AI companies must find ways to monetize their most valuable users while managing the costs of serving free users who help improve the product.