Communications of the ACM
How to Predict the Progress of Technology
Researchers considering models that forecast the pace of technological progress found that Moore's Law and Wright's Laws offer the best predictions.
Credit: Casey Fleser/Wikipedia
Among existing models to forecast the pace of technological progress, Moore’s Law and Wright’s Law offer the best predictions, according to Massachusetts Institute of Technology and Santa Fe Institute researchers.
Their research shows the best predictor of technological progress is Wright’s Law, written in 1936, which states that every percent increase in cumulative production results in a fixed increase in production efficiency.
The second best predictor of technological progress, the researchers found, is Moore’s Law, written in 1965, which states that components in integrated circuits will double every 18 months, or more generally that improvement rates increase exponentially over time, with different rates for different technologies.
The researchers collected information for at least a 10-year time period on actual costs and production levels for 62 industry sectors, and analyzed the data to see which formulas best matched the rate of technological progress.
Rates of change vary widely for different technologies, with information technology advancing very rapidly.
The research could enable regulators to forecast economic effects of policy, help businesses focus their research efforts, and enable investors to determine high-growth sectors.
From MIT News
View Full Article
Abstracts Copyright © 2013 Information Inc., Bethesda, Maryland, USA
No entries found