A very loud version of the AI conversation right now is about replacement: which jobs go, which professions disappear, how many humans will be unnecessary in five years. It is a story about subtraction. It also happens to be the wrong story.
The technology revolutions that produced the most enduring companies were not the ones that subtracted people. They were the ones that made people more capable. Spreadsheets did not replace accountants — they turned every analyst into someone who could model an entire business. Search did not replace librarians — it gave anyone with a question the answers a research department used to provide. The companies built around those shifts compounded for decades.
Augmentation, not replacement
The most useful frame for evaluating an AI company is not "what does it replace" but "what does it make possible." The companies we want to build, back, and acquire are the ones whose customers finish the day having done something they couldn't have done at all the day before — not the same thing, faster.
A solo founder running an entire back office. A first-time entrepreneur launching a real business in a weekend. A small team serving customers in twelve languages without hiring twelve people. A student learning a subject at the level a private tutor used to provide. Those are the shapes of human potential being unlocked. Those are the companies worth building.
The test we apply
When a company comes through our process, we ask three questions:
- Who becomes more capable? If the answer is unclear, we are not interested.
- What were they unable to do before? New capability is more durable than incremental efficiency.
- Does the product treat them as agents or as inputs? We back companies that put the human in charge.
The longer view
Markets that empower people compound. They generate loyal customers, organic distribution, and the kind of word-of-mouth growth that no paid acquisition strategy can replace. Markets that merely extract efficiency from people produce a different kind of company — leaner in the short run, weaker in the long run.
We are not in this work to subtract humans from the economy. We are in it to multiply what humans can do. The companies built on that idea will be the ones that matter twenty years from now.