Video · 2026-07-14 · 52m · 6 moments

The Trillion-Dollar Industries AI Is Disrupting: Voice, Law & the End of the Billable Hour

✦ AI generated

timeline · colored by role

Data

ElevenLabs grew from launching its first product in early 2023 to $600 million in ARR in under four years, with growth accelerating each round: 20 months to $100M, 10 months to $200M, and just 5 months to $300M.

ElevenLabs' CEO details a revenue ramp that kept accelerating, going from launch in 2023 to $600M ARR in under four years, with each $100M-plus milestone reached faster than the last.

transcript

Mati Staniszewski: We started company 2022. First year was all about building the the research and the product to really kickstart the work. We built the first texttospech model that finally could sound human. Released it in 2023, beginning of 2023. Then it took us roughly 20 months to get to the first 100 million in ARR. Roughly 10 months to get to 200, 5 months to get to 300.

Fact

ElevenLabs built a marketplace where people can license their own cloned voice and get paid for it, and has already paid back over $22 million to that community of voice talent.

In response to concerns about voice cloning and impersonation, Mati describes ElevenLabs' authenticated voice marketplace that has paid out over $22 million to voice talent, alongside safeguards like tracing and moderation.

transcript

Mati Staniszewski: As a company, what we now do and this from the beginning, we created a marketplace where people can create their voice. We authenticated, you can share it, and you earn money. Today, we paid back over $22 million back to the community of of of talent.

Mechanism

ElevenLabs stays ahead of Anthropic and OpenAI in voice not through model scale but through specialized architecture, proprietary labeled audio data from over a thousand contractors, and deep vertical product work that generalist labs won't do.

Facing competition from Anthropic and OpenAI, ElevenLabs' CEO says their edge comes from research architecture and a large in-house data-labeling operation, not from scale, plus vertical product depth those labs won't replicate.

transcript

Mati Staniszewski: It's the architecture that matters, not the scale. You really need to change how the model operates. Two, you need very specific data that there's of course a wide set of data out there, but it's unlabelled data and where we spend a lot of time. So we build a internal team of over thousand contractors that label all those audio assets to make them good.

gives example · 1

Prediction◆

AI voice agents have already passed the Turing test for conversation, and within the year any voice conversation will feel indistinguishable from talking to another human.

Mati Staniszewski argues that text-based AI has effectively passed the Turing test and predicts voice AI will reach the same milestone within the year, making conversations indistinguishable from talking to a human.

transcript

Mati Staniszewski: It's it's almost crazy that each of us has like a a Turing uh like you know if you if you were to chat with an agent now it feels like um the Turing test will be complete it's as smart as another human. Uh and we hope this year we'll do that same thing for voice uh where any conversation feels like you are speaking with another human.

supports · 2

Data◆

The legal services industry is a trillion-dollar-a-year market that remains almost entirely manual and fragmented, with software making up only about 4% of spend against 96% for human services.

Legora's founder frames legal as a trillion-dollar annual market that is overwhelmingly delivered manually, with legal technology capturing only a sliver (~4%) of total spend — implying massive room for AI-driven software growth.

transcript

Max: you have this enormous bucket of legal services which today is being done manually. It's a trillion dollars every year into legal services which is very fragmented.

Claim

The billable hour business model at law firms only survives by overcharging junior associates' hourly rates to compensate for underpricing the much higher value of senior partners' time, and AI is starting to break that model.

Max argues the traditional billable-hour law firm model is really a cross-subsidy scheme — overcharging associates and undercharging partners — and that AI tools are starting to break this structure by automating associate-level work.

transcript

Max: it it starts to um break this model where you charge out associates for very high hourly rates and you have a billable hour model. And actually, if you look in law firms, the way that that business model works is you overcharge for the associates and you actually undercharge for the partners.

rebuts · 1

Highlight slides

Text AI Has Already Passed the Turing Test✦ from: AI voice agents have already passed the Turing test for conversation, and within the year any voice conversation will feel indistinguishable from talking to another human.Voice AI: The Next Turing Milestone✦ from: AI voice agents have already passed the Turing test for conversation, and within the year any voice conversation will feel indistinguishable from talking to another human.Legal: A Trillion-Dollar Manual Market✦ from: The legal services industry is a trillion-dollar-a-year market that remains almost entirely manual and fragmented, with software making up only about 4% of spend against 96% for human services.The Opportunity for AI✦ from: The legal services industry is a trillion-dollar-a-year market that remains almost entirely manual and fragmented, with software making up only about 4% of spend against 96% for human services.

Related episodes

Video · The Cognitive Revolution

Intelligence on the Edge: Liquid AI's Ramin Hasani on the Search for Device-Native Foundation Models

→

Article · Import AI

Import AI 465: Open vs closed gaps; Kimi K3; Demis' big policy plan

→

Video · a16z Podcast

Garry Tan: New Rules for Founders

→