The AI EA Flex

$1.6 Billion of Equity On-Chain. Here Is Why.

Picture St. Louis, 1849. Two men with a bottle of champagne and a brutal choice: go north for beaver pelts, or go south chasing gold in California. Joris Delanoue does not even blink. He picks the gold. Not for the metal. For the belief. That single line tells you everything about this conversation. Joris, co-founder and co-CEO of Fairmint, has spent the better part of two decades pushing into frontiers nobody else wanted to settle. He sold a cloud computing company, Nexteem, before most people trusted the cloud. Now he is doing the same thing to the one document that quietly governs every startup's destiny: the cap table. Here is the uncomfortable truth he lays out. Fifty years ago, Microsoft went public at an eight hundred million dollar valuation, and ordinary people built entire retirements on the climb that followed. Today, companies stay private until they are worth five hundred billion, and the upside goes to a happy few. The frontier did not close. It just moved behind a velvet rope. Joris wants to tear the rope down, and he thinks the tool to do it is equity that lives on a blockchain: programmable, transferable, and liquid enough that the engineer who bet fifteen years of her life on a startup can actually borrow against her shares to buy a house. This is a conversation about wealth, about who gets to build it, and about why the most defensible thing you own in the AI era is no longer your product. It is your distribution. Buckle up. Compliance by Automation (not Intermediation)Joris frames the entire Fairmint thesis as a shift away from people and toward code. * Old world: compliance happens through layers of intermediaries, lawyers, banks, and reconciliation. * New world: compliance lives inside a smart contract that mimics securities law and applies the rules automatically. * The promise: lower transfer costs, fewer trolls under the bridge, and a single source of truth for who owns what. The Shovel Seller's DilemmaThe gold rush metaphor that opens the episode is a strategy lesson in disguise. * The prospector grinds sixteen hours a day and often ends with broken backs and empty pans. * The shovel seller monetizes everyone else's dream regardless of who strikes gold. * Joris flips it: do not just sell shovels, own a piece of the mine through programmable equity. https://www.fairmint.com/ https://www.linkedin.com/in/delanoue/ https://www.linkedin.com/in/estesryan/⁠⁠ ⁠⁠https://aiforfounders.co⁠⁠ https://trynina.co/ https://ainativestudent.com/

Tiny Brands Are Outranking Billion-Dollar Companies. Here Is the Hack.

Everything you know about getting found online is about to be obsolete. For two decades, founders chased one algorithm. Backlinks, page speed, keyword stuffing, the whole exhausting machine. Then a handful of chatbots quietly took the wheel, and now they decide which businesses get recommended and which ones get ghosted. SEO is bleeding out. Google traffic is leaking. And a huge slice of buyers now ask ChatGPT to pick their solution before a human ever enters the conversation. Here is the twist that should make every founder sit up straight: the playing field is wide open for the first time in twenty years. Tiny brands are outranking billion-dollar incumbents. Unknown podcasts are beating the giants. The new game is not about who has the biggest backlink pile. It is about who tells the clearest story. In this return visit, Jenna Hannon, co-founder and CMO of Hatter, runs a live tactical teardown using Ryan's own podcast page as the guinea pig. She walks through what AI search actually rewards, why a lead who found you through ChatGPT shows up on the call already sold, and the exact content structure that gets your brand cited instead of buried. She also drops the unglamorous truth about where AI pulls its recommendations from, and it is probably not where you think. This is your chance to control your narrative before the bots write it for you. The Three Names, One Thing PrincipleThe category has a branding problem. AEO, GEO, AI search. They sound different. They are not. * AEO stands for answer engine optimization. * GEO stands for generative engine optimization. * AI search is the plain-language version. * All three describe the same goal: showing up inside AI chatbots. Ignore the LinkedIn posts insisting they are separate disciplines. The Magical MomentThe new tell that AI search is working for you. * A lead arrives having done their research inside a chatbot, not a Google rabbit hole. * They already know they have a problem and that you are one of the few recommended solutions. * They land on the call warm, informed, and close to buying. * The chatbot's endorsement carries trust, because the user already treats the AI like a trusted advisor. https://gethatter.ai [https://gethatter.ai] https://aiforfounders.co [https://aiforfounders.co] https://www.linkedin.com/in/jennahannon [https://www.linkedin.com/in/jennahannon] ⁠⁠https://www.linkedin.com/in/estesryan/⁠⁠ https://trynina.co/

The Founder Is the Bottleneck. Here's How to Clone Your Judgment.

You are the smartest person in your company. That is exactly the problem. Every founder hits the same wall. The strategy lives in your head. The taste lives in your gut. The thousand tiny judgment calls that make your company yours live nowhere anyone else can reach them. So your team waits. They wait on your approval, your context, your answer to a question you have answered nine times already. And while they wait, the work does not move. Joshua Liberson, CEO and co-founder of Dobbin, has spent a career watching this play out. He designed editorial systems for magazines, ran brand and creative at One Kings Lane, and advised a long list of founder-led companies before deciding the bottleneck was always the same: the founder cannot be in every room. Dobbin is his answer. It is a company AI that captures the fifteen-or-so dimensions of an organization, its culture, values, brand, strategy, and objectives, and then delivers that judgment to every person on the team right inside Slack, where the work already happens. The pitch is deceptively calm. Dobbin is not a creative generator and not a design tool. It is a thinking partner. The designer drops a layout into a channel and Dobbin critiques it against the principles the team itself articulated. The intern asks what to do today. The CEO uses it for high-value strategic thinking. Josh's favorite proof point is a creative agency built around the photographer Mark Seliger, whose Dobbin was assembled from four and a half hours of audio about a forty-five-year career in lighting, composition, and printmaking. The result: a managing director who now answers RFPs in thirty minutes instead of three weeks and seventeen meetings. Underneath the warm language is a hard claim about modern work. Microsoft estimates 57% of our time goes to coordination, roughly 22 hours of a 40-hour week. Nobody's KPI is "coordinate more," yet that is what the calendar quietly becomes. Josh's fix is not more project management, which he thinks the world already drowns in. It is what his friend Howard calls ambient alignment: the strategy is simply present, in the channel, evolving as the company evolves, so people stop waiting and start shipping. And he is honest about the banana peels. A great team is a pirate ship, full of brilliant misfits who wither under too much rigidity. So Dobbin is built to bend. It is iterative, never bedrock. It watches where work drifts from the foundation, then proposes amendments the founder can accept or reject. Structure that empowers, not structure that scolds. Or, as Josh puts it through a borrowed line from a Greek philosopher, you never step in the same river twice, because the river is flowing and so are you. LinksDobbin: https://dobbin.ai [https://dobbin.ai] Joshua Liberson on LinkedIn: https://www.linkedin.com/in/joshliberson/ AI for Founders newsletter: https://aiforfounders.co [https://aiforfounders.co] Ryan Estes on LinkedIn: https://www.linkedin.com/in/estesryan/ [https://www.linkedin.com/in/estesryan/]

Your Face, Voice, and Data Are Fakeable. Here's What Isn't.

Everything you trust online is about to break, and André Ferraz built a company to catch the people breaking it. Picture a 12-year-old kid riding his bike through Brazil when a stranger points a gun at his face to steal it. That kid grew up with two computer scientist parents, an early love of code, and a peculiar fascination not with building systems but with breaking them. Three decades later, that instinct for thinking like an attacker became the foundation of Incognia, a company now embedded in 1.2 billion monthly active devices and built on a single contrarian belief: your location behavior is the strongest signal of who you really are. But the road there nearly ended before it began. André moved to the United States six years ago to chase the biggest market, bringing a thriving location-based advertising business with him. Then the pandemic hit. Physical retailers shut down. Revenue collapsed 95% in a single month. The team went from 250 people to 50, keeping only the engineers. Most founders would have folded. André and his co-founders looked at the precise location technology they had spent over a decade perfecting and asked a different question: what else can this do? The answer was fraud prevention, and it turned out the world needed it desperately. Incognia now serves banks, fintechs, crypto exchanges, and marketplaces, answering one deceptively simple question for every login, transaction, and signup: is this user who they say they are? The results speak loudly. Triple revenue growth. Six times the return on investment delivered to clients. A 100% trial-to-paid conversion rate. And a 180% net dollar retention rate that means customers keep expanding once they see the data. The conversation gets genuinely unsettling when André lays out the asymmetry of modern fraud. The criminals are professionals, not hoodie-wearing loners. They run 60,000 fake accounts in two days. They factory-reset devices in 30 seconds to dodge detection. They share tools and open-source software while the banks defending against them compete and stay siloed. The money pouring into making deepfakes dwarfs the money fighting them. As André puts it, if you brought him a deepfake detection company, he would not invest, because detection can never outspend generation. So Incognia plays a different game entirely. Rather than analyzing whether a video is a deepfake, it checks whether the camera feeding that video is even real. Rather than trusting a spoofable GPS coordinate, it fuses Wi-Fi, Bluetooth, cell tower, compass, accelerometer, and gyroscope signals to locate a device down to eight foot accuracy, close enough to separate two fraudsters in different apartments of the same building. The bet is that AI can fake your face, your voice, and your data, but it cannot cheaply fake the real physical world at scale. Make the attack economically unfeasible, and the fraudster moves on. The episode closes on a vision that goes beyond catching criminals. André imagines a world without buttons, where the hotel TV logs you in automatically, the thermostat already knows your preferred temperature, and the world quietly personalizes itself around you because it recognizes you everywhere. Stop the fraud first, because that hurts. Then make the world more elegant. The Asymmetry of FraudAndré's core mental model for why defenders are structurally disadvantaged: * Criminals break rules freely while banks must follow heavy financial and privacy regulation. * Fraudsters collaborate and share open-source tools while competing banks stay siloed. * Deepfake generation attracts vastly more capital than deepfake detection ever will. * The takeaway: never fight on the attacker's terms, find a different angle. https://www.incognia.com/ [https://www.incognia.com/] https://www.linkedin.com/in/andreferraz/ [https://www.linkedin.com/in/andreferraz/] ⁠⁠https://www.linkedin.com/in/estesryan/⁠⁠ ⁠⁠https://aiforfounders.co⁠⁠ https://trynina.co/

Data agents you can trust in production. | Pradnesh Patil from Altimate AI

Pradnesh Patil spent years as a product leader at Fortune 500 companies bringing in millions in revenue, and every single quarter the same thing kept happening. He would walk into leadership meetings, present data, get hit with the question "why is this number different from last week," and then watch the opportunity window close while his data team spent two months trying to figure it out. It was not a bad data team. It was every data team. Data work is complicated, the experts get pulled in fifteen directions, and the backlog never shrinks. So Pradnesh called up Aaron, his co-founder of ten years and a veteran data and ML engineering leader, and they did the same thing they had done a dozen times before: they built something together. Last time it was an autonomous crypto trading bot. This time it was Altimate AI, a company built on a single insight. The bottleneck in enterprise is not engineering talent. It is the gap between the institutional knowledge locked inside your 15 year veteran's head and the new hires who do not have it yet. They raised on a few slides without writing a line of code, then built a free product that hit a million downloads across 100+ countries. That feedback flywheel turned into an enterprise offering, a second funding round, and Fortune 500 logos. The latest chapter is Altimate Core, an open source agent data engineering harness that now sits at number one on the industry benchmark. The Four Components of an Agent Harness * Context: Metadata pulled from across the hybrid data stack, plus the tribal knowledge previously locked in employee heads. * Governance: Rules, permissions, and access controls that respect regulated industries like healthcare and financial services. * Tools and Skills: The specific recipes and connectors agents need for specialized data work. * Infrastructure: Sandbox environments for hundreds of agents to work in parallel without touching production. The Tribal Knowledge Capture Loop * The system watches a senior engineer fix a problem and stores how they did it. * When a less experienced person hits the same issue, the system recalls the fix and recommends it. * Users can correct the memory when AI picks up the wrong pattern. * Active coaching of agents is positioned as the new responsibility for senior engineers. The Token Efficiency Stack * Route reasoning heavy tasks like data modeling to frontier models. * Route simple tasks like writing column descriptions to cheaper models. * Bring your own LLM, including open source, to control costs and meet governance requirements. * Avoid brute forcing one model into every specialized task. The Four High Value Data Use Cases * ELT pipeline development and debugging. * Data infrastructure optimization, with cost reductions of 30 to 40 percent. * Governance reporting and sensitive data tracking. * Legacy stack migrations without paying a services firm millions. https://www.altimate.ai/ [https://www.altimate.ai/] https://www.linkedin.com/in/pradneshpatil/ [https://www.linkedin.com/in/pradneshpatil/] https://www.linkedin.com/in/estesryan/⁠⁠ https://www.hssv.org [https://www.hssv.org/] ⁠⁠https://aiforfounders.co⁠⁠ https://www.youtube.com/@AIforfounders1