DTIM 2026: The fastest technology always wins

The puzzle is growing faster than anyone can lay down the pieces. There are now more than 265,000 technology areas to watch, and no R&D team can hire its way to covering them. “From this point forward, it is going to get exponentially worse.”

Your experts and your dashboards both get it wrong

Internal experts know the technologies you already build, not the ones coming to replace them. “It is very difficult to find a battery guy that likes hydrogen, or the other way around.” Trend tools have the opposite flaw: they count. Filings, funding rounds, startups. Counting finds what is popular, and “just because a lot of people are doing something does not mean you should. It just means that they are.”

The law nobody taught you

MIT spent 30 years measuring how 28 technology domains improve. Every one improves exponentially, at a steady rate. Only the speed differs: optical telecoms double their performance every 14 months, milling machines take 23 years. Two numbers hidden in patent citation data predict that speed before the market sees it. Cycle time is how fast a field produces its next invention. Knowledge flow is how heavily each invention gets built on.

Batteries vs. hydrogen, called in 2001

In 2001, lithium-ion batteries were improving at about 55% a year. Hydrogen fuel cells managed about 21%. The car industry spent the next two decades betting on hydrogen anyway. The same gap showed up between solid-state and hard drives, and between online and offline payments. Every time, the faster improver won.