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

Original Media · 52m · English · Podcast · All-In with Chamath, Jason, Sacks & Friedberg

TLDR

This episode features two interviews with AI company CEOs: Mati Staniszewski of ElevenLabs (voice AI) and Max Junestrand of Legora (legal AI). ElevenLabs reached $600M ARR in 40 months with 600 employees by embedding engineers in every function and eliminating product managers, while pioneering voice AI that consumers now prefer over humans for speed and emotional safety. Legora is transforming the $1T legal services market (only 4% software) by providing AI-native tools that enable in-house legal work, cross-jurisdictional research, and agent-based workflows, growing 50% quarter-over-quarter for seven quarters.

For you

PRO members can sign in and request more information — or their own version of this digest, tuned to what they care about.

Sign in to unlock

Surprising Insights

•

ElevenLabs never had product managers; they embed engineers in non-engineering teams (legal, talent, sales) to build internal tools and enforce AI security.

•

People are more honest with AI voice agents about sensitive topics (e.g., debt) because they feel less shame than with human operators.

•

Legora's 'legal engineers' (forward-deployed lawyers) sit with law firm partners to transform their business models from billable hours to fixed fees.

•

The US legal research market is a duopoly (Westlaw/LexisNexis) where case law is privately owned, not public, requiring physical scanning of court records.

•

ElevenLabs' research team outcompetes frontier labs on voice models because 'architecture matters more than scale' and they invest heavily in labeled audio data (1000+ contractors).

Key takeaways

•

ElevenLabs achieved $600M ARR in ~40 months with 600 employees, maintaining culture through research-product integration and embedded engineers across all teams (legal, talent, sales).

•

ElevenLabs operates without product managers; engineers are embedded in every function to drive AI adoption, build internal tools, and enforce security checks on AI-generated code.

•

Voice AI has crossed a quality threshold where consumers prefer AI agents over humans for speed, lack of social friction, multilingual capabilities, and emotional safety (e.g., discussing debt without shame).

•

ElevenLabs implements safeguards (traceability, moderation, detection) and a voice marketplace paying creators $22M+, while enabling voice restoration for ALS/throat cancer patients.

•

Legora is transforming the $1T legal services market by providing AI-native tools that enable in-house legal work, cross-jurisdictional research, and agent-based workflows, growing 50% QoQ for 7 quarters.

Notable quotes

“We built the first text-to-speech 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 million. Five months to get to $300 million. And that's how we closed end of the last year. And now we are at $600 million.”

▸ Notable quotes · 1:07

“We don't have any PMs. Never did.”

▸ Notable quotes · 7:14

“With AI, people are much more open to share what actually happened, give the information. And suddenly this emotional block of like, in front of other human, I don't want to be able to say all of that is very different.”

▸ Notable quotes · 15:00

“I don't believe in fine tuning or building any general intelligence models. I think that's a total waste of time and money. I do believe in very narrow models for narrow use cases that you also drive a lot of scaling so you can drive both cost and latency down.”

▸ Notable quotes · 48:51

Chapters

•

0:42 — ElevenLabs: Revenue Growth & Organizational Structure: Mati describes ElevenLabs' rapid revenue ramp from $0 to $600M ARR in 40 months, the company's 600-person headcount, and their unique research-product integration model with zero attrition among the first 10 hires.

•

4:22 — ElevenLabs: No Product Managers & Embedded Engineers: ElevenLabs operates without product managers; instead, engineers are embedded in every function (legal, talent, sales) to build automations, drive AI adoption, and enforce security checks on AI-generated code.

•

9:58 — ElevenLabs: Voice AI Quality & Consumer Preference: Voice AI has crossed a quality threshold where consumers prefer AI agents for speed, lack of social friction, multilingual support, and emotional safety — people are more honest with AI about sensitive topics like debt.

•

15:37 — ElevenLabs: Safeguards, Celebrity Voices & Voice Restoration: ElevenLabs implements three safeguards (traceability, moderation, detection), runs a voice marketplace paying creators $22M+, partners with celebrities like Matthew McConaughey, and restores voices for ALS/throat cancer patients.

•

25:45 — ElevenLabs: Competition with Frontier Models: ElevenLabs stays model-agnostic, outcompetes frontier labs on voice quality through architecture innovation and labeled data (1000+ contractors), and is developing its own narrow models for interaction-specific tasks.

•

31:44 — Legora: Transforming the $1T Legal Market: Max describes Legora's attack on the $1T legal services market (only 4% software), enabling startups to replace lawyers with AI for incorporation, cap tables, HR, and due diligence, while law firms adopt fixed-fee models.

•

37:47 — Legora: Law Firm Partnerships & Legal Engineers: Legora deploys 'legal engineers' (forward-deployed lawyers) to help firms like Kirkland transform from billable-hour to fixed-fee models; Legora itself closed an acquisition in 12 days using its own tool.

•

40:47 — Legora: Cross-Jurisdictional Data & Competing with LexisNexis: Legora structures global case law and legislation to enable cross-border legal work; the US duopoly (Westlaw/LexisNexis) owns case law privately, but Legora's data moat and agentic workflows are disrupting them.

•

47:35 — Legora: Narrow Models vs General Intelligence & Data Security: Max rejects building general legal intelligence models, focusing instead on narrow fine-tuned models for specific tasks (e.g., tabular contract review); Legora hosts national secrets and weapons contracts without on-prem deployment.

Speakers

•

Jason Calacanis — Host / Investor

•

Mati Staniszewski — CEO, ElevenLabs

•

Max Junestrand — CEO, Legora

⋮