Where this term comes fromIn this channel's episode with Nabeel Hyatt (Spark Capital), Japanese toilets of AI is one of three types of innovation he borrows from Rory Sutherland to explain what he hunts for as an investor: - Faster horses — incremental improvement. Most venture funding flows here, and most of it gets “arbitraged out of existence” within four years. - Teleportation — the grand obvious dream, like Cruise and self-driving cars. Everyone wants it; nobody knows how to get there. - Japanese toilets — the thing you didn't know you needed until you experienced it, then wonder why it isn't everywhere. Not even that complicated — but the execution is subtle and hard. How Nabeel applies itHe says: “I am looking for mostly new experiences — things that surprise you with the place that they're playing.” His canonical example is Granola, an AI notetaker that looks like Apple Notes, quietly appends extra context to notes you were already taking, and makes you a little smarter. Once you play with it, “if you have any taste or product sense,” you know it's the thing. Granola's first version was a misfire — an interruptive tab-completion tool. Nabeel passed at seed. Only months later, when the founder stumbled into the current form, did Spark move quickly, and a ~$20M round followed. Why it mattersFor Nabeel, Japanese toilets of AI is the category worth backing: subtle, delightful, hard-to-execute products rather than obvious bets. The founder trait behind them is sensitivity — close listening to real behavior, paired with an orientation to movement. - Japanese toilets of AI names the innovation type Nabeel Hyatt hunts for: subtle, hard-to-execute products that surprise you with delight once you experience them. - It comes from Rory Sutherland's three types of innovation, contrasted with faster horses (incremental, mostly arbitraged away) and teleportation (grand obvious dreams). - Granola is the canonical example; sensitivity is the founder trait behind such products.