This browser is not actively supported anymore. For the best passle experience, we strongly recommend you upgrade your browser.
| 3 minutes read

Are papers and patents becoming less disruptive over time?

In January, Nature published an article on the disruptiveness of papers and patents over time. In this study they took data from 45 million papers and 3.9 million patents, and, using a new quantitative index called the CD index, determined whether innovation is progressing or slowing. The CD index is a citation-based measure ranging from −1 (consolidating) to 1 (disruptive), with the assumptions that if a paper or patent is disruptive, the subsequent work that cites it is less likely to also cite its predecessors, or instead if a paper or patent is consolidating, subsequent work that cites it is also more likely to cite its predecessors.

From the viewpoint of a patent attorney, I have some questions with the patent aspect of the study, and believe that there could be several factors at play here that probably wouldn’t equally effect the scientific paper data. Authors of scientific papers are more likely to include a large bank of citations, explaining all the previous research that has helped them arrive at their discovery. In contrast, whilst there are some patent office requirements to cite all relevant documents, it can be advantageous to limit citations in a patent and so authors might not look for any more references than they aren’t already aware of.

The article comes to the conclusion innovation is becoming less disruptive and notes that “For both papers and patents, the rates of decline are greatest in the earlier parts of the time series, and for patents, they appear to begin stabilizing between the years 2000 and 2005”.

An uptick in disruptiveness starts to be seen from around 2005 onwards, especially in the technology areas of computers and communications, and electrical and electronics. However, the data only extends to 2010, and so does not provide any insight into the disruptiveness of the Fourth Industrial Revolution.

For a number of reasons, it would be particularly interesting to see how this trend has developed beyond 2010. Since 2010 software and electronics, in particular artificial intelligence, has been increasingly implemented in other fields. In addition to this, as 3D printing has become more viable, new innovative products are less limited by manufacturing methods. It is not inconceivable that, were the study to be repeated with more recent data, the 4th Industrial Revolution would coincide with a period of increased disruptiveness.

On the other hand, since 2010 consumer electronics have risen in popularity. At the same time, the speed of the internet has increased, and users demand more and more content from their devices. A relatively small increase in battery life or display refresh rate can still be very beneficial to users. This, in combination with the cycle of releasing new products every year, could mean that applicants are filing patent applications for less disruptive technology with smaller differentials over existing technology.

AI inventors

In March 2023, the UK Supreme Court is set to hear an appeal on two UK patent applications listing DABUS (Device for the Autonomous Bootstrapping of Unified Sentience) as the inventor, to decide whether an AI-based machine should be allowed to be regarded as an inventor for purposes of the UK Patents Act. This case will have important implications, as AI systems are increasing being used in a number of fields such as drug discovery. This could lead to a further increase in the disruptiveness of new technology, though if patents can’t be filed for inventions devised by AI this wouldn’t be reflected in any repeat of the study!

Titles and Abstracts

In the article, Nature also did some textual analysis of titles and abstracts, assuming that disruptive papers and patents are likely to introduce new words. This also found a decline in the disruptiveness of patents, though this could be due to reasons relating to patent law itself rather than the innovation. The title of a patent is often published before the full specification – this can lead to purposefully ambiguous titles being used when applicants want to keep the focus of their innovation secret. Abstracts of patents generally use similar language to the independent claims and in order to avoid terms in the claims being misconstrued, patent attorneys will often avoid using new terms in the claims. Instead, it is common to use broad terms that encompass a larger scope than just the invention itself where possible, and only use narrower - and likely newer - language where it has a clearly defined meaning.  

Across fields, we find that science and technology are becoming less disruptive... For both papers and patents, the rates of decline are greatest in the earlier parts of the time series, and for patents, they appear to begin stabilizing between the years 2000 and 2005.


3d printing, 4ir, artificial intelligence, digital transformation, internet of things, patents, robotics