Technology and the Economy: What is Actually Happening?

Humans are excellent storytellers, and we’re inundated with stories about technology, companies, and stocks. Sometimes these stories paint an unbiased picture, but not always. Instead, let’s look at the data and see what stories it tells.

Methodology: 

  1. Collect all the US patents in a rolling 3-year window. 
  2. Count the number of patents issued each day in each of the 255,374 unique CPC subclasses. 
  3. Sort the subclasses by their sizes and growth rates to observe where the most technology innovation is happening, and where there is recent growth. 
  4. Count the patents per owner/assignee within these fast growing domains, and sort them.

Here we see the fastest growing areas of innovation, with the largest volumes of new patents, and recent acceleration where new activity in the last 30 days stands out.

H10D30/01 Manufacture or treatment

Some classifications have experienced rapid growth of new patent issuance over the last 4 years. Some highlights include: Field-effect transistors [FET]

H10K59/80 Constructional details

H10D30/67 Thin-film transistors [TFT]

H10D84/01 Manufacture or treatment

H10D62/10 Shapes, relative sizes or dispositions of the regions of the semiconductor bodies; Shapes of the semiconductor bodies

H10D84/03 using Group IV technology, e.g. silicon technology or silicon-carbide [SiC] technology

A61K40/42 Cancer antigens

H04L41/16 using machine learning or artificial intelligence

C12N15/86 Viral vectors

H10D64/01 Manufacture or treatment

H10F39/00 Integrated devices, or assemblies of multiple devices, comprising at least one element covered by group H10F30/00, e.g. radiation detectors comprising photodiode arrays

Notes and Disclaimers: Data includes US Patents from 2017 to the present (updated 2x/week). Data is estimated to be more than 99.95% complete, but a small number of very long patents (e.g. including DNA sequences) failed to properly save. In this document we make casual use of the terms “CPC Classifications”, “classifications”, “classes”, “domains”, “markets”, etc. and we have shortened or modified some classification names in order to make the ideas clearer. Please ask if you would like clarification about anything. Because we are counting the patents within each classification, patents with multiple classifications will show up within the each, so readers should not treat the counts as additive when evaluating higher level classifications; counts must be recalculated at the parent class in order to avoid double counting.