A patent is only a licence to litigate

Vikram's claim is that intellectual property is not a trophy but an instrument of business strategy, and most deep-science founders hold it the wrong way round. Patents are only one of two buckets, he argues; the other is know-how — the IP you deliberately chose not to file, and the kind that leaves on two legs when people do. A patent, in his phrase, is only a licence to litigate. What makes it valuable is structure: US continuations let you add claims in the three years before allowance, once you finally know which vertical is fertile, and Indian divisionals split one filing into four patents that can each be licensed to a partner holding the capital you don't have — worth more, he says, than a field-of-use carve-out on your master patent. The second half of the session is harder, because deep tech satisfies a need that nobody yet wants. His answers are counterintuitive and lived: let two established players copy you, because an incumbent copying a startup makes the market for it; open the books when you fail, because that is where brand equity comes from; and when the customer's disbelief is the only obstacle left, sell them insurance against your own failure — which for a year and a half made premiums a high double-digit share of EBIT, until customers noticed nothing was breaking and simply bought more product. The stakes are the sector's: more than half the world's iron-making plants are only half amortised, so carbon-free steel has to work inside the existing mill before it can replace it, and the founder who cannot size that pilot has to rent the nose that can.