David Sacks | All In Summary

1. Key Themes

ElevenLabs' Hypergrowth Trajectory Is a Case Study in AI-Native Scaling

ElevenLabs went from zero to $600M ARR in roughly 40 months, with an accelerating velocity: 20 months to $100M, 10 months to $200M, five months to $300M, and then doubling to $600M from there. This is one of the fastest revenue ramps in enterprise software history.

"We started the company 2022. First year was all about building the research and the product to really kickstart the work. 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. Five months to get to 300. And that's how we closed end of the last year, and now we are at 600." 00:01:00

Legal Tech Is the Most Underinvested Trillion-Dollar Market in Software

The legal services market is $1 trillion annually, yet software penetration is only ~4% ($40B), meaning 96% of legal work is still done manually. The speaker from Legora frames this as the single largest software-to-services conversion opportunity in any industry.

"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. But the software spend into legal technology is about 40 billion. So it means there's 4% software, 96% service, which is bananas." 00:34:09

The Billable Hour Model Is Structurally Broken and AI Will Collapse It

Law firms have historically overcharged on associate hours to cross-subsidize partner time. AI breaks this model by eliminating the need for armies of associates doing rote work — forcing a shift to fixed-fee, success-fee, and outcome-based pricing.

"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... The only way they know how to price that is to overcharge for the associates." 00:35:31

Voice Is Becoming the Default Interface — Consumers Are Actively Requesting AI Agents

The ElevenLabs CEO describes a tipping point where consumers now prefer talking to AI agents over humans, and enterprises are seeing a "golden era" unlock in voice-first customer journeys — from reactive support to proactive, personalized interaction.

"We are seeing a transition where suddenly... the product combines the reliability that's core with the orchestration for a lot of the AI models... And yeah, I think it was a step change in the last 12 months and especially in the last six of how good that experience became. Where it's like this golden era for the consumers out there." 00:09:58

AI Is Enabling Vertical Specialization Over Horizontal General Intelligence

Both ElevenLabs and Legora explicitly reject building general-purpose AI models and instead double down on vertical-specific data, workflows, and integrations. This is a deliberate strategic choice with strong economic rationale.

"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." 00:48:49

Legacy Data Moats Are Weaker Than Previously Assumed

The conventional wisdom was that companies like LexisNexis and Westlaw, sitting on decades of proprietary legal data, would be the AI era winners. The Legora CEO directly disputes this, noting that their stocks are being crushed and that AI-native startups can replicate or partner around the data advantage faster than legacy players can transform.

"At the outset of AI, many believed and made a bet that those organizations who had all the data was going to be the winners. As we're starting to see in the market, that's no longer the case." 00:43:52

Embedding Engineers Into Every Function — Including Non-Technical Teams — Is a Structural Competitive Advantage

ElevenLabs embeds engineers inside legal, talent, and go-to-market teams. These engineers have a dual role: building automation and serving as a security/quality gate on AI-generated code deployed by non-engineers.

"Our talent team will have an engineer. Our legal team will have an engineer. Our revenue engineering or go-to-market engineering have engineers embedded all across. And those people have two roles. One is, of course, creating automations and bringing the software inside of that team. But second is actually helping everybody else do what you said, which is make sure that people are adopting AI, but also there's a security check for everything they deploy." 00:05:05

The Compliance Moat in Legal AI Is Harder to Cross Than the Technology Moat

Legora's CEO argues that building legal AI is technically straightforward — the real barrier is getting through enterprise compliance and security requirements. Once inside, expansion economics are highly favorable, and this also drives their M&A strategy.

"There's a lot of legal AI companies and very few are making it through. And not because it's hard to build stuff. It's actually quite easy to understand where you can build value. But getting into the customer is very hard. But that's something we cracked pretty early on. And once you're in, it's much easier to expand. So that's also one of the driving forces behind our M&A strategy." 00:50:01

Voice as Identity and IP Is an Emerging Asset Class

ElevenLabs has paid over $22 million back to voice talent through its marketplace, and is working with celebrities (Matthew McConaughey, the James Earl Jones estate, Jamie Foxx) to monetize voice IP across languages and interactive formats. This is a new and growing economic category.

"Today we paid back over $22 million back to the community of talent... voice is such a big part of identity and probably our most important work was actually working with people that lost their voice due to ALS, due to throat cancer and working on bringing that voice back." 00:20:53

2. Contrarian Perspectives

Legacy Data Holders Will Not Win the AI Era — Incumbency Is a Liability, Not an Asset

Conventional wisdom says whoever owns the data wins. The Legora CEO directly contradicts this: Westlaw and LexisNexis are watching their stocks decline despite sitting on monopolistic legal data assets. The bottleneck has shifted from data to execution tempo, talent, and product architecture.

"I think at the outset of AI, many believed and made a bet that those organizations who had all the data was going to be the winners. As we're starting to see in the market, that's no longer the case... They can't get the talent. They don't work our hours. And they're so political in their organizations that it's just hard to move." 00:43:52

Claude Entering Legal Is a Pipeline Generator for Legora, Not a Threat

While most observers would see Anthropic launching a legal AI product as existential competition, the Legora CEO views it as a top-of-funnel demand generator — customers try Claude's shallow legal offering, hit the ceiling, and call Legora.

"Claude has a legal offering, which is basically a bundling of markdown skills files and a couple of integrations. And so I think what's really helpful about that is that it illustrates to everyone how applicable AI is in law... What it also does is it drives a lot of initial usage there, and then you hit the ceiling. Or you understand how shallow it is, and then you call us. So it's actually a big pipeline generator for us." 00:47:47

No Product Managers Is Not a Cost-Cutting Measure — It Is the Right Architecture

ElevenLabs has never had a single product manager, and the CEO frames this as an intentional, principled structural decision, not a scrappy startup compromise. The reasoning: AI now allows domain experts in one field to perform competently in adjacent fields, eliminating the need for a coordination layer.

"We don't have any PMs... I thought it's a little bit of what you mentioned... the ideal person in that role can code, can understand the customer, can understand design... What we are seeing now, if you can do a little bit of all with AI, you can maybe step change from being an amateur to being an advanced level, maybe not an expert level. So suddenly you are not bottlenecked on all the other functions to do your work." 00:07:13

People Are More Honest with AI Than with Humans — and This Is a Feature, Not a Risk

In financial services debt collection contexts, customers share more truthful information about their financial situations with AI voice agents than with human agents because there is no social shame. This is counterintuitive and has direct business model implications.

"Frequently people would naturally feel ashamed of telling the real situation. With AI, people are much more open to share what actually happened, give them advice, information. And suddenly this emotional block of like, in front of another human, I don't want to be able to say all of that, is very different." 00:14:34

The U.S. Legal Data Infrastructure Is Privately Owned — and That Is a Structural Moat and Systemic Risk

Westlaw holds what is effectively a government-granted monopoly on U.S. case reporting. The data is not publicly owned, which creates both a structural barrier to new entrants and a meaningful societal inefficiency that AI-native companies must physically work around.

"Westlaw basically has a monopoly with the American government to report on the cases. So they're not owned by the public in a way. They're owned by a company... You cannot build a legal research solution that doesn't have all of the data. Because if you go to Wachtel and a litigator at Wachtel says, I'm going to use this to go after Elon or do a billion dollar case, you better make sure you have all the cases." 00:45:11

3. Companies Identified

ElevenLabs

AI voice platform specializing in text-to-speech, speech-to-text, voice cloning, and agent orchestration. Built from scratch in 2022, now at $600M ARR with 600 employees, powering voice experiences for enterprises in telco, financial services, healthcare, gaming, and entertainment. Partnered with Epic Games (Fortnite/Darth Vader), MasterClass, Headspace, and celebrity voice IP deals.

"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. Five months to get to 300. And that's how we closed end of the last year, and now we are at 600." 00:01:00

Legora

AI-native legal platform serving law firms and enterprise legal departments. Achieved 50% quarter-over-quarter growth for seven consecutive quarters, recently crossing $150M ARR (described as one of the fastest enterprise direct-sales companies to reach that milestone). Conducting M&A diligence in-house with their own tool, completing transactions in 12 days from LOI to close.

"We actually just became, as of the close last week on Tuesday, one of the fastest enterprise company with a direct sales motion to go from one to 150, beating Sierra with one quarter." 00:32:00

Harvey

Legal AI company referenced as a key competitor to Legora in the U.S. market. Named by Jason Calacanis as Legora's most prominent American contemporary.

"With your tools, obviously you got your contemporary and Harvey and people... And then you also have, I guess, Claude and other folks also want to be in your business." 00:32:13

Airwallex

AI-native global financial platform for accounts, cards, and payments. Described as built for the intelligent era from day one rather than bolting AI onto legacy infrastructure.

"If you were building a global financial system from first principles today, you wouldn't build it on 50-year-old legacy rails. You'd build Airwallex, one AI-native platform for global accounts, cards, and payments." 00:00:00

Cooley

Top-tier law firm noted for proactively building a software platform to serve startup founders directly, embedding their legal precedent and contract review workflows. Cited as an example of a legacy player adapting its business model under AI pressure.

"Cooley actually started serving startup founders directly with a sort of software platform. But you just log onto the platform. They've pumped it full with their material and their precedent. And then you have the startup material there. And they've embedded workflows that reviews the contracts." 00:35:02

Kirkland & Ellis

Top U.S. law firm generating $10 billion a year in revenue with 4,000-5,000 lawyers, with partners earning $5-10M in annual profits each. Cited as the epitome of the high-margin law firm model most threatened by AI disruption.

"Kirkland Ellis turns around $10 billion a year... per partner, they make between 5 and 10 million every year in profits." 00:37:58

Wachtell Lipton

Cited as the best law firm in the world for litigation, used to illustrate the zero-tolerance-for-error standard in legal data completeness for high-stakes cases.

"If you go to Wachtell and a litigator at Wachtell, the best law firm in the world, says, I'm going to use this to go after Elon or do a billion dollar case, you better make sure you have all the cases." 00:45:42

Westlaw (Thomson Reuters)

Dominant U.S. legal research database with a government-granted monopoly on case reporting. Cited as a legacy incumbent under existential pressure from AI-native competitors, with declining stock price as evidence.

"Westlaw basically has a monopoly with the American government to report on the cases. So they're not owned by the public in a way. They're owned by a company." 00:45:11

LexisNexis (RELX)

The other half of the U.S. legal research duopoly. Despite making a couple of billion dollars annually and holding massive case law data, its stock is declining amid AI disruption.

"LexisNexis has been a juggernaut and the legacy player in all the case law and regulations. They have a massive data moat. But they only make a couple of billion dollars a year... are they coming for our business?" 00:42:31

Whisperflow

Voice-to-text productivity tool used by Jason Calacanis, powered in part by ElevenLabs. Enables continuous stream-of-consciousness prompting via pedal interface.

"They use us and a few others as well. They are doing phenomenal work too. Whisperflow is just a tremendous product." 00:12:01

Plaud

Wearable AI recording device mentioned approvingly by the ElevenLabs CEO for capturing signal from in-person conversations and events.

"I have the Plaud. It's incredible. Plaud, pocket, phenomenal. Like so good. And especially in events like this... how incredible would it be that all the signal on the conversations that otherwise disappear, you maybe tap few notes here and there to try to get signal afterwards." 00:12:26

Epic Games / Fortnite

Gaming company that partnered with ElevenLabs, Disney, and the James Earl Jones estate to create a live interactive Darth Vader voice agent within Fortnite gameplay.

"Fortnite launched Darth Vader, which people and players could interact with live in partnership with the estate, in partnership with Disney. So every player after reaching a certain stage could have a Darth Vader interact and help you solve the missions." 00:23:26

MasterClass

Online education platform working with ElevenLabs to transform static celebrity-taught content into interactive, voice-powered personalized learning experiences.

"They worked with talent directly. And here you have previously a static content that you would learn from. Now you have interactive content. So you have Gordon Ramsay teaching you how to cook in the kitchen. He can scream at you if you're not doing it right." 00:20:22

Headspace

Meditation app using ElevenLabs for content localization and interactive, personalized meditation experiences.

"Headspace has a great meditation... they are localizing a lot of the content." 00:23:55

Calm

Meditation app; Jason Calacanis disclosed he is an investor since the company was valued at $4M. Noted as experimenting with interactive, personalized meditation elements.

"Could you have a meditation lesson that's personalized to you? Which we would hopefully love to... I did invest in Calm, but it was a $4 million company. But Calm is incredible." 00:24:34

Revolut

Fintech company; ElevenLabs customer in the financial services vertical.

"We work with a lot of financial services companies, Revolut, Klarna, PagBank." 00:14:34

Klarna

Fintech/BNPL company; ElevenLabs customer, including for payment reminder and debt collection voice agent use cases.

"We work with a lot of financial services companies, Revolut, Klarna, PagBank." 00:14:34

PagBank

Brazilian fintech; ElevenLabs customer in financial services.

"We work with a lot of financial services companies, Revolut, Klarna, PagBank." 00:14:34

Oracle Cloud Infrastructure

Mentioned as the infrastructure partner for AI companies training and deploying at scale.

"The AI companies building the future run on Oracle Cloud Infrastructure, training and deploying at scale on one of the world's largest AI infrastructures." 00:31:07

Palantir

Referenced as the model for Legora's "forward-deployed lawyer" concept, analogous to Palantir's forward-deployed engineers embedded with customers.

"In the same way that Palantir has forward-deployed engineers, we have forward-deployed lawyers. And their job is to sit down with the Kirkland partners and help them transform their business from a pre-AI to a post-AI world." 00:38:48

Sierra (AI company)

Referenced as the prior record-holder for fastest enterprise direct-sales company to reach $150M ARR, a record Legora claims to have beaten by one quarter.

"We actually just became, as of the close last week on Tuesday, one of the fastest enterprise company with a direct sales motion to go from one to 150, beating Sierra with one quarter." 00:32:00

4. People Identified

Mati Staniszewski (ElevenLabs CEO/Co-founder)

Co-founder and CEO of ElevenLabs. Credited with assembling a zero-attrition research team and scaling to $600M ARR. Spoke about company architecture, voice AI research strategy, and competitive positioning against OpenAI and Anthropic.

"From the original team, very research, very engineering heavy. From the first 10 people, we had zero attrition. Everybody is still at the company from those core research and engineering talent building together with us." 00:02:58

Max Junestrand (Legora CEO/Co-founder)

Co-founder and CEO of Legora. Scaled the company to 50% quarter-over-quarter growth for seven consecutive quarters. Previously acquired four businesses in a single year, closing the fastest transaction in 12 days. Speaks with deep structural knowledge of the legal data landscape.

"We acquired four businesses so far this year. We did the diligence in-house with our own tool. And the fastest transaction we did was 12 days from LOI to closing." 00:36:32

Matthew McConaughey

Hollywood actor; partnered with ElevenLabs for a multilingual voice IP deal — the first voice licensed to carry across languages with full emotional fidelity.

"We partnered with Matthew McConaughey on creating a world. All right. All right. All right. And across languages. And as the first... the crazy thing with what AI technology opens is that now the voice can be carried not only in English, but also in Spanish and Italian and Portuguese." 00:19:17

James Earl Jones

Iconic voice actor (Darth Vader). His estate partnered with Disney and ElevenLabs to power the interactive Darth Vader voice in Fortnite, enabling real-time player interaction at scale.

"Before he passed, I think he did a deal with Disney and he said, for my family, I would like to license the Darth Vader voice for all time to Disney... instead they went to you." 00:22:41

Jamie Foxx

Actor; mentioned as a celebrity who has been paid for their voice through ElevenLabs' voice IP licensing program.

"I think you got Jamie Foxx and some other folks actually that you paid for their voices." 00:18:15

Jennifer Wexton

U.S. Congresswoman who lost her voice to a degenerative condition and used ElevenLabs to deliver what was described as the first AI-restored voice used in a congressional speech.

"We worked with Congresswoman Jennifer Wexton, who lost her voice and wanted to continue to inspire others that you can do incredible work despite that. And was the first speech delivered in Congress." 00:21:57

Gordon Ramsay

Celebrity chef; his voice and teaching persona are used in MasterClass interactive content powered by ElevenLabs, where he can respond to students in real time.

"You have Gordon Ramsay teaching you how to cook in the kitchen. He can scream at you if you're not doing it right. Scallops are raw." 00:20:22

Sergey Brin

Google co-founder; referenced for his advice to "threaten LLMs with bodily harm" as an effective prompting technique.

"We saw Sergey Brin say threaten it with bodily harm. It's a very effective technique if you haven't tried it." 00:14:21

5. Operating Insights

Use AI-Powered Voice Intake to Replace Web Forms — It Captures More Qualified Information

ElevenLabs deployed an inbound AI SDR agent that lets prospects call and speak instead of filling out a form. The insight: spoken intake produces richer, more contextual information than typed forms, leading to faster routing to the right person and better conversion.

"In addition to the form that you fill on the website, you have an agent that you can call. And people are, of course, can give all the information in a much easier and quicker way. But the second thing that happens is people also leave a lot more information so you can get connected to the right problem and right person a lot quicker." 00:08:34

Do M&A Diligence In-House With Your Own AI Tool to Compress Deal Timelines Dramatically

Legora used its own legal AI platform to conduct diligence on four acquisitions, closing one transaction in 12 days from LOI to close. This is a direct demonstration of how founder-led, AI-native diligence eliminates the incentive misalignment of external counsel (who benefit from deals taking longer).

"We acquired four businesses so far this year. We did the diligence in-house with our own tool. And the fastest transaction we did was 12 days from LOI to closing." 00:36:32

Build Small, Tightly-Knit, Industry-Specialized Teams Rather Than Functional Monoliths

ElevenLabs operates in five-to-ten person pods, each optimized for a specific vertical (telco, financial services, healthcare) and each containing embedded engineers. This prevents the bureaucratic drag of large functional departments and keeps decision velocity high.

"We do that across the company. So it's usually five to ten people teams that run ahead. And inside of each of those teams, the decision we took, which is slightly different than how it's usually structured, we embedded engineers in every place." 00:05:05

Pedal-Activated Continuous Voice Prompting Unlocks a New Level of LLM Output Quality

Jason Calacanis describes a hardware-software workflow — a foot pedal paired with a voice-to-text tool — that enables one-to-two minute continuous stream-of-consciousness prompts. This exploits LLMs' strength at synthesizing large, unstructured verbal input and removes the friction of typing that causes people to under-prompt.

"When I press the pedal down, I just give a stream of consciousness now. And it turns out what these LLMs actually do really well with is taking a massive stream of consciousness where you just keep talking and talking and talking. So I'll give it a one to two minute prompt. Then I let go. And it has changed everything." 00:13:25

6. Overlooked Insights

The Trillion-Dollar Legal Market Will Be Won Through Physical Data Acquisition — Not Just Software

Buried in the legal discussion is a deeply non-obvious structural reality: to serve high-stakes legal clients at all, you must have 100% of case law — not 80%, not 95%. And in the U.S., that data is locked behind a Westlaw government-granted monopoly that requires physically shipping books, opening them, and scanning them page by page for citation accuracy. This means the true moat in legal AI is not the AI — it is the willingness to do unglamorous, expensive, physical data acquisition work at scale. Any AI-native legal company that skips this step is locked out of the most lucrative litigation clients forever.

"You cannot build a legal research solution that doesn't have all of the data. Because if you go to Wachtell and a litigator at Wachtell says, I'm going to use this to go after Elon or do a billion dollar case, you better make sure you have all the cases... You have to physically get the books all the way to India. You need to open them. You need to scan them because you need to get what's called page citations." 00:45:42

Anthropic's Claude Opus 4.5/4.6 Represents a Qualitative Leap in Agentic Legal Reasoning — Not Just Incremental Improvement

The Legora CEO briefly mentions "especially the agents following the release of Opus 4.5 and 4.6" as the specific catalyst enabling AI to move from legal research augmentation to genuine end-to-end case strategy — combining witness statements, case law, and litigation strategy autonomously. This throwaway line signals that a specific model version triggered a step-change in what legal AI can actually do, and that the transition from "AI helps lawyers" to "AI does legal work" may already be underway right now, not in some future state.

"What's really interesting about especially the agents following the release of Opus 4.5 and 4.6 is they can now start to do really intelligent case strategy. And they can actually start to combine the witness statements, the cases, and they can really do end-to-end work, which is, I think, moving us from a world where AI is just augmenting to AI is actually really doing things." 00:46:46