What is the best way to predict a technology cost decline. The [Santa Fe RIsk Institute](https://sfi-edu.s3.amazonaws.com/sfi-edu/production/uploads/sfi-com/dev/uploads/filer/c2/a5/c2a50ab2-0efb-4742-86ce-7065938c40c8/12-07-008.pdf) did a thorough analysis to answer this questions. In 2012 the researchers compared the forecast error rates between Moore’s Law and Wright’s Law across 62 technologies ranging from black and white TVs to photovoltaic cells, and from electric ovens to nuclear power. They concluded that [[Wright’s Law]] seems to outperform [[Moore's Law]] consistently Even in research on transistors, Gordon Moore would have been served better by applying [[Wright’s Law]] than creating [[Moore's Law]] as a forecasting tool, as shown below.   Measured over the decade to 2015 (the latest data point available), a price forecast based on Wright’s Law was **40% more accurate** than one based on Moore’s Law. Measured across the entire data series Wright’s Law reduced the forecast error by **15% on average**. ![[Pasted image 20210919215935.png]] #concept