General-purpose technology

In economics, it is theorized that initial adoption of a new GPT within an economy may, before improving productivity, actually decrease it due to time required for development of new infrastructure, learning costs and, obsolescence of old technologies and skills. This can lead to a "productivity J-curve" as unmeasured intangible assets are built up and then harvested. The impending timeframe to utilize the latent benefits of the new technology is deemed a trade-off. Spin-out firms/inventors from organizations that had developed GPTs play an important role in developing applications. However, it has been observed that the level of cumulative innovation in GPTs diminishes as more spin-outs into application development occur.

Historical GPT

Economists Richard Lipsey and Kenneth Carlaw classified 24 technologies in history as true GPTs. They define a transforming GPT according to the four criteria listed below:

1. is a single, recognisable generic technology
2. initially has much scope for improvement but comes to be widely used across the economy
3. has many different uses
4. creates many spillover effects

Since their book, more GPTs have been added for the 21st century.

Foundational

The earliest technologies mentioned by Lipsey and Carlaw occur before the Neolithic period and have not been cast as GPTs, however, they are innovations that the other 24 rely upon.

Expanded list of 26 technologies

Steam engine increased labor productivity annually by 0.34%; IT by 0.6% (1995–2005); robotics by 0.36% (1993–2007).