Hank’s investment claim: spending real capital — time and money, not just attention — on an idea that probably will not work is one of the highest-return moves available, because the failure itself returns tools the toolkit did not previously contain.
Verbatim: “at various times in my life I have said okay it is all right for me to spend 5% of my net worth on this idea that probably won’t work and when those things have failed in my life which early on they all did I came out with new tools.”
The point is not the ideas. Hank corrects himself mid-sentence: “I feel like those were really smart ideas and sorry they were not smart ideas they were bad ideas.” The bad ideas worked anyway — not as products, but as toolkit-builders. Working with another person, learning a new domain under pressure, hitting and naming the actual blockers (focus, cost, marketing, story, distribution) — that is what the failed project taught.
He distinguishes this from two safer ways to spend on growth: (1) buying education from a university, (2) investing or lending to someone else. Both are legitimate but produce a different kind of learning. Self-funded execution forces you to discover what was actually in the way — “Was it me? Was it my ability to stay focused on the task? Was it the cost getting bigger than I thought it was going to? The marketing being harder than I thought it was going to. The story not working for people…” The diagnostic only runs when your own capital is on the line and the project is yours.
Why this beats “fail fast” in the abstract. Hank’s frame is not “iterate quickly through bad ideas.” It is “spend enough that the failure is real.” 5% of net worth is large enough that you will care, small enough not to be ruinous. The capital threshold is the point — under it, you are not actually testing the toolkit, you are pretending to.
The trap to avoid. This is not a license to romanticize failure. Hank flags his own rose-colored glasses about the era. The claim is narrower: capital spent on shipping a real attempt — even one that fails — returns more toolkit than the same capital spent on credentials or on funding someone else’s attempt. The mechanism is forced confrontation with delivery, not failure as a virtue. Compare 2026-04-26-creating-a-solution-is-easier-than-delivering-it — most ideas die at delivery, and only your own failed deliveries teach you what delivery actually requires.
Mentoring application. For someone deciding between safe credential-building (another course, another certification) and a self-funded experiment likely to fail, the toolkit-return on the experiment is usually higher — if enough capital is committed to make the attempt real. The path most people take is neither: they take the course because it is safe, and dabble in the experiment without enough commitment to learn from its failure.