My First Million · Episode Brief
7 Business Ideas from a Unicorn Founder
Immad Akhund is running a unicorn bank and daydreaming about nuclear power plants on the moon, and somehow both feel like the same person.
Immad Akhund founded Mercury, which has become one of the most-used banking products in the startup ecosystem by doing something ambitious and obvious simultaneously: building a bank that treats founders like founders rather than like small business owners. This conversation is Shaan asking Immad what he would build if Mercury didn't exist — a productive hypothetical that reveals how his mind categorizes opportunity.
The lunar nuclear power plant idea opens the episode in deliberately absurd territory, but Immad defends it with the same reasoning pattern he applies to everything: what is the scale of the problem, what is the timing of the enabling technology, and what is the structural reason that no one has solved it yet? In the case of lunar nuclear power, the answers are 'enormous,' 'approaching,' and 'nobody has been willing to take the regulatory and capital risk' — which is a description that applies to a lot of opportunities Immad finds interesting.
The Peter Thiel test segment is the episode's most transferable idea: Thiel's framing that every startup should be able to answer the question 'what important truth do very few people agree with you on?' as a filter for contrarian insight. Immad uses this as a quality check for his own business ideas — if the answer is something most people already agree with, it's not the right entry point. The 'enterprise agents' idea is his most immediately commercializable pitch: AI agents that handle specific, high-value business workflows (not general-purpose assistants) with enough context about a specific company to be genuinely useful rather than generically capable.
The 'find accepted abuse' segment is where Immad is most directly useful as a product thinker. His definition: accepted abuse is any workflow, product, or service that people tolerate as 'just the way things work' despite being objectively bad — and every successful startup he's seen has exploited one of these. Mercury itself is built on the accepted abuse of traditional business banking. His advice for finding your own version is to look for categories where the incumbent NPS score is negative and ask why nobody has fixed it yet.
Key Ideas
- →The Peter Thiel test as a business idea filter: ideas are only worth pursuing if they're built on a belief that very few people currently hold — if your thesis is consensus, the opportunity is already being competed for.
- →Enterprise agents (narrow AI agents designed for specific high-value business workflows with deep company context) are more defensible than general-purpose assistants because the switching cost compounds with each workflow they embed into.
- →The 'digital twin CEO' concept: an AI agent trained on a founder's communication patterns, decision history, and strategic priorities that can handle the repetitive decision-making load that consumes executive bandwidth.
- →'Find accepted abuse' as the Mercury origin story: banking for startups was objectively bad (slow account opening, no API access, terrible UX), and the fact that founders tolerated it was evidence that the TAM was real and the switch cost was low.
- →Asymmetric bets deserve asymmetric commitments: Immad's framework for how to allocate time and capital to long-horizon bets (like nuclear power) versus near-term plays (like enterprise agents) without being paralyzed by opportunity cost.
Worth Remembering
Immad explaining why he'd build a lunar nuclear power plant with the same flat affect he uses to explain Mercury's banking product — the indifference to absurdity is the tell.
The 'selling stars' idea: literally selling naming rights to stars as a gift product, which Immad treats seriously as a brand-building exercise because the margins on naming certificates are extraordinary.
The Silicon Valley Bank failure discussion, where Immad as a bank founder has more direct standing to comment than most podcast guests — and says things about the structural fragility of the system that most people in banking won't say publicly.
Shaan asking 'choose your mess' at the end — the idea that every company has an inescapable dysfunction, and the founder's job is to choose the mess they're most capable of tolerating.