Processing Artificial Intelligence: Introduction
From: Canadian Intellectual Property Office
The rapid pace of innovation in Artificial Intelligence (AI) technologies, and the fact that it touches many industry sectors, makes this a topic that is top of mind to a wide range of stakeholders. Although the main advances in AI have been developed over the past seventy years, the discussion around the creation of an alternative intelligence dates back to the 1300s.Footnote 3 By 1943, the idea of "artificial neurons" spurs the interest in computer-based neural networks and then deep learning.Footnote 4 In 1950, Alan Turing proposes what is now fondly called the "Turing Test": an imitation game that tests a machine’s intelligence by imitating the sentient behaviour of a human.Footnote 5 This sequence of events ultimately led to the term "Artificial Intelligence" first being coined in 1955 by John McCarthy, Marvin Minsky, Nathaniel Rochester and Claude Shannon in a research proposal that looked to explore the thinking capacity of machines.Footnote 6
The significant progress made in the early days of AI was followed by what is now commonly known as "AI Winter": a period of time when ". . . commercial and scientific activities in AI declined dramatically".Footnote 7 The decline ended in the 1970s and greatly stalled any further progress in AI.Footnote 8 Despite this, there has been relentless progress in current years. The availability of data in combination with higher computing power has caused an explosion in the field.Footnote 9
Establishing a universally accepted definition of AI from a patent perspective is a challenging feat. Due to the ever-evolving nature of AI, its patent definition needs to be continuously adjusted by incorporating the latest terminology to capture breakthroughs. The World Intellectual Property Organization (WIPO) was one of the first Intellectual Property (IP) institutions to provide a general overview of patenting in AI using a patent-based definition. Subsequently, the United Kingdom Intellectual Property Office (UKIPO) developed its own search strategy and released a report articulating the AI patenting landscape in the United Kingdom (U.K.). Building on the work of the UKIPO, CIPO has adopted the methodology used by our colleagues in this report and expanded it to present the patent landscape from the perspective of Canadian inventions filed in Canada and abroad rather than focusing on those only filed at CIPO (additional details in Annex A). The intent of this research is to highlight the areas where Canadian researchers and institutions are most innovative and identify where they may be relatively more specialized globally. Examining Canadian researchers and institutions separately provides a deeper understanding of the state of innovation in AI. Understanding Canada’s technological strengths from the perspective of researchers and institutions is helpful to policymakers when developing targeted policies that can be designed to increase our strength in specific technology fields with the ultimate objective of advancing innovation.
Ultimately, this report is the culmination of a literature review, a compilation of organization-specific information, an extensive and targeted patent search strategy, and a close examination of the differences between the filing activity of Canadian institutions and those of Canadian researchers, regardless of their affiliations to institutions and the nationality of those entities. The next section of the report discusses the patent dataset used as an indicator for innovation in the AI field. The third section presents an overview of the international patent landscape based on the origin of the names assigned to the patented inventions, hereafter referred to as assignees. The fourth and fifth sections look at AI inventions patented by Canadian institutions and Canadian researchers, respectively. These two sections provide a detailed overview of filing activity looking at areas of specialization by AI sub-category, key players, geographical distribution across the country, and patent landscape maps. The details presented in these sections are useful to better understand the evolution and the current state of innovation in this technology field. The sixth section follows CIPO’s collaboration with Statistics Canada and sheds light on the industry, size, and ownership characteristics of the Canadian institutions patenting in AI.
- Date modified: