Exponential Technology Forecasting
Forecast adoption through growth curves and people building the future
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 6
- Confidence
- 93%
Exponential Technology Forecasting begins by separating human linear intuition from the measured curve of an enabling technology. Identify the capability that drives progress, such as computing power or gene-sequencing throughput, and determine whether it is improving exponentially. Then interview scientists and engineers doing the work, because their plans already incorporate technical milestones and constraints that outsiders miss. Convert those inputs into timelines for applications, while checking whether nontechnical bottlenecks could interrupt the curve. Kaku explains that seemingly implausible predictions later arrived because computing power followed Moore's law and practitioners understood that compounding. The method does not assume every technology grows forever; it requires repeated measurement and revision when the curve or its constraints change.
Origin
Michio Kaku explained that he forecast technologies by interviewing hundreds of working scientists and accounting for exponential computing growth rather than relying on ordinary linear intuition.
Core principles
- 01Human intuition defaults to linear change
- 02Some enabling technologies improve exponentially
- 03Builders use different timelines from outside observers
- 04Forecast the enabling curve before predicting applications
How to run it
- 1
Find the enabling curve
Identify the measurable capability or cost whose improvement makes the target technology possible.
Pro tip Look beneath the visible product to the infrastructure that controls its feasibility.
Watch out Do not assume a trend is exponential because recent progress feels fast.
- 2
Measure the growth pattern
Use historical performance or cost data to distinguish linear, exponential, and plateauing change.
Pro tip Plot repeated measurements rather than extrapolating from two points.
- 3
Interview active builders
Ask practitioners what they can do now, what milestone comes next, and which bottlenecks remain.
Pro tip Prefer people running experiments over commentators describing the field.
Watch out Builders can still be optimistic about their own work, so compare several independent views.
- 4
Translate capability into applications
Estimate when the enabling curve crosses the threshold required for a concrete product, experiment, or use case.
Pro tip State the threshold explicitly so the forecast can be updated.
- 5
Check external bottlenecks
Test whether regulation, manufacturing, biology, cost, or adoption could delay the application even if the technical curve holds.
Pro tip Separate technical feasibility from widespread deployment.
Watch out A fast component does not make the entire system exponential.
- 6
Refresh the forecast
Re-measure the curve and update practitioner inputs as milestones arrive or fail.
Pro tip Record which assumption changed rather than silently moving the date.
In the wild
A scientist challenged Kaku's sequencing forecast based on the painful speed of processing one gene. Kaku points out that the work nevertheless accelerated until the entire genome was sequenced, and the cost later fell to a few hundred dollars.
→ Exponential improvement overturned a forecast based on the field's initial linear pace.
For Physics of the Future, Kaku interviewed roughly 300 leading scientists across computing, robotics, biotechnology, and space travel. He says practitioners' timelines differ because they understand the exponential computing curve driving their work.
→ Builder interviews and an enabling growth curve supplied a grounded basis for long-range technology forecasts.
Common mistakes
Projecting a straight line
Linear intuition can badly underestimate a capability that compounds at a stable exponential rate.
Interviewing only commentators
People actively building the technology have more direct evidence about milestones and constraints.
Ignoring deployment friction
Technical capability can grow rapidly while regulation, production, economics, or adoption delays widespread use.
Is it for you?
Best for
It is best for fields where a measurable capability such as computing or sequencing cost drives downstream applications.
Not ideal for
It is not ideal when adoption is dominated by regulation, scarce physical inputs, social resistance, or a growth curve that has already flattened.
From the transcript
“I interview the people who are working on these Technologies they're well aware of Morris law and as a consequence their time frame is much…”
“our brain is linear it is not exponential but what drives a lot of these Technologies is computer power which obeys Mo's law which is…”
From the episode
#562: Dr. Michio Kaku — Exploring Time Travel, the Beauty of Physics, Parallel Universes, the Mind of God, String Theory, Lessons from Einstein, and More
Michio Kaku