# The First AI Market With 8 Billion Potential Users | Sudarshan kamath, Smallest AI

Page: https://stenobird.com/podcast/the-neon-show-862453/the-first-ai-market-with-8-billion-potential-users-sudarshan-kamath-smallest-ai
Text version: https://stenobird.com/podcast/the-neon-show-862453/the-first-ai-market-with-8-billion-potential-users-sudarshan-kamath-smallest-ai.md
Podcast: [The Neon Show](https://stenobird.com/podcast/the-neon-show-862453)
Published: 2026-03-06T11:30:00+00:00
Episode link: https://www.buzzsprout.com/1601695/episodes/18800879-the-first-ai-market-with-8-billion-potential-users-sudarshan-kamath-smallest-ai.mp3
Audio file: https://www.buzzsprout.com/1601695/episodes/18800879-the-first-ai-market-with-8-billion-potential-users-sudarshan-kamath-smallest-ai.mp3
Processing state: not_requested
JSON: https://stenobird.com/v1/public/podcasts/the-neon-show-862453/episodes/the-first-ai-market-with-8-billion-potential-users-sudarshan-kamath-smallest-ai
Duration seconds: 4165

## Resource

Will smaller AI models win over large language models? Sudarshan Kamath grew up in Mumbai, taught himself AI before most Indian companies were even hiring for it, and bought the domain "smallest.ai" for $100 in 2022, two years before the company existed. Today, he runs Smallest AI, a startup focused on real time voice AI. He started with self-driving cars, training large models and compressing them to run on vehicle hardware in real time. That's where he first saw what small models could do...

## Actions

- request_transcript: `POST https://stenobird.com/v1/public/podcasts/the-neon-show-862453/episodes/the-first-ai-market-with-8-billion-potential-users-sudarshan-kamath-smallest-ai/transcription-requests` — Idempotently request low-priority transcript generation for this episode.
- read_markdown: `GET https://stenobird.com/podcast/the-neon-show-862453/the-first-ai-market-with-8-billion-potential-users-sudarshan-kamath-smallest-ai.md` — Read the agent-friendly Markdown representation of this episode resource.

A page view does not enqueue transcription. Agents should invoke `request_transcript` explicitly when they need this episode processed.

## Transcript

Full transcripts are not published on public pages unless there is a clear rights basis.