Human beings think linearly. Technology moves exponentially. I have lived through several platform shifts, and timing is still the lesson I have mastered least. Exponential change takes longer at the beginning than anyone expects. Then it happens much faster at the end.
Walk across a room. One step, two steps, three steps. Every step is the same size. We evolved in that world. Technology does not live there. It compounds. Ray Kurzweil filled a book with charts of it. I remember paging through them and thinking they all looked the same.
Put a linear mind in front of an exponential curve and it makes the same mistake every time. It draws a straight line through the slow part and concludes that nothing much is happening.
An analysis of the history of technology shows that technological change is exponential, contrary to the common-sense “intuitive linear” view.
1994: inside the slow part.
When Joel Maske and I launched NETworth, there were no stock quotes anywhere on the internet. HTML had no way to make a table, so we could not line up the columns of numbers. When we showed it to people who understood markets, some dismissed it. The web was a toy, they said. Serious investors would never trust a browser.
They were drawing a straight line through the slow part. People told me they would never put their account information on the internet. Within a decade nearly everyone did. We were so early that, in hindsight, I could have held on and become ETrade before ETrade.
1999: I made the other mistake.
The futurist Roy Amara put the consequence in one sentence I quote in nearly every lecture. We tend to overestimate the effect of a technology in the short run and underestimate it in the long run. I have lived both halves. In 1999 I was on the wrong side of the first.
At Internet Venture Works we built online versions of businesses owned by buyout firms. The idea of shipping someone a heavy bag of dog food through the mail was almost absurd. The vision was right. The timing was wrong. When the market broke in 2000, we could not hold on long enough for the world we saw to arrive. It arrived anyway, on other vehicles. Within two decades, Chewy built a large public company selling pet food online.
2019: early, with staying power.
The second time I was early, I made sure we could stay in the game. In 2019 Ciaran Hynes, our team at COSIMO and I launched COSIMO X, a tokenized evergreen venture fund. Plenty of smart people told us we were early. They were right. We built to outlast the wait. If you are early, the crawl lasts longer than your plan says. Fund the time it will take.
Now add AI.
AI is not one more technology on Kurzweil’s charts. It speeds up the others. Most leadership teams I meet make both of Amara’s mistakes at once. They expect this quarter’s pilot to transform the company. They underinvest in the slow work that matters over five years: redesigning roles, training people, building the data and rules the systems will need. A leader who budgets in straight lines is not just late. Late arrives faster than it used to.
Where is it on its curve? And am I drawing a straight line through the slow part?
Read the exponent.
- Name the curve under youWrite the one technology your business depends on most. Find how fast its cost or capability has compounded over the last five years.
- Place itIs it in the slow, flat start or past the turn? Write the evidence, not the feeling.
- Check your plan for straight linesLook at hiring, capital and runway. Mark every number that assumes steady, linear growth.
- Fund the crawlIf you are early, write how long you can stand at the door. If the answer is shorter than the crawl, fix that first.
- Plan for the turnWrite what you will do if the curve turns next year and moves faster than your organization can follow.
Quick answers.
What does it mean to think in exponents?
Rob Frasca uses it to mean asking two questions about any technology: where is it on its curve, and am I drawing a straight line through the slow part? Linear intuition underestimates compounding technology until the curve turns.
Why do good founders get the timing wrong?
Rob Frasca argues that exponential change takes longer at the beginning than anyone expects and happens much faster at the end. He was early with NETworth in 1994, too early with Internet Venture Works in 1999, and early again with tokenization at COSIMO, where he built for staying power.
Sources.
- Rob Frasca, The Creator’s Journey (forthcoming), Chapter VI, Seeing the Shift.
- Ray Kurzweil, “The Law of Accelerating Returns,” essay, 2001. Wording checked against the essay as reproduced by the Lifeboat Foundation. Source
- On the origin and wording of Amara’s law: Quote Investigator, “We Overestimate the Short-Term and Underestimate the Long-Term,” 2019. Source