Can voice AI finally pass for human? ElevenLabs’ co-founder explains the gap
This preview, based only on the published show notes, points to a conversation about why voice AI remains harder than text and where ElevenLabs sees the biggest opportunities. Expect discussion of audio models, translation, voice agents, and the company’s rapid rise.
This episode preview is based on the published show notes, so think of it as a guide to what the conversation appears to cover rather than a recap of the audio itself. From the notes, John speaks with Mati Staniszewski, co-founder of ElevenLabs, about the current state of voice AI and why conversational speech is still a tougher challenge than text. A central thread seems to be the idea of a "voice Turing Test" or conversational Turing Test: not just whether AI can generate speech, but whether it can hold up in natural back-and-forth interaction. The episode also looks set to get practical. The timestamps suggest a walk through how audio models work, the tradeoffs between cascaded systems and speech-to-speech approaches, and the possibility of universal translation. If you're curious about where voice products may actually land in the real world, the notes point to use cases ranging from farming to healthcare, plus a discussion of why seemingly simple tasks like having a phone read a PDF still fall short. On the company side, there appears to be a business and organizational angle too: ElevenLabs’ model, its rapid climb to an $11 billion valuation, and what it means to design an AI-native organization. The mention of Ukraine using the company’s technology for digital government services adds a policy and public-sector dimension that could make this especially interesting for listeners beyond pure AI research. If you want a founder-level view of where voice AI is promising, limited, and commercially relevant, this looks like a strong pick.
About this episode
<p>Mati Staniszewski is the co-founder of ElevenLabs, the research company making audio accessible across languages and voices. He sits down with John to discuss the "voice Turing Test" and why AI has conquered text but still struggles with conversational speech. They discuss the future of human-computer interaction, including why we still can't get our phones to read a PDF properly and the massive potential for voice agents in everything from farming to healthcare. Mati also opens up about ElevenLabs’ rapid ascent to an $11 billion valuation and gives a behind-the-scenes look at how Ukraine is using their tech for digital government services.</p><p><br></p><p>Timestamps</p><p>(00:00:27) How audio models work</p><p>(00:08:52) ElevenLabs business model</p><p>(00:17:50) The conversational Turing Test</p><p>(00:21:01) Link by Stripe</p><p>(00:26:02) Cascaded vs speech-to-speech</p><p>(00:31:53) Universal translation</p><p>(00:51:41) Designing an AI-native org</p>