From the course: Introducing AI to Your Organization

Why bother introducing AI to an organization?

From the course: Introducing AI to Your Organization

Start my 1-month free trial

Why bother introducing AI to an organization?

“

- [Instructor] So why bother introducing AI into your organization? So I thought instead of providing hype from consultancies, I would give you data that is openly available and then you get to decide for yourself. Patents provide a valuable means of assessing trends and research, as they reveal the areas of innovation that inventors are focused on. By analyzing patent data, it is possible to track changes over time and identify which jurisdictions are seeing most patenting activity. The World Intellectual Property Organization 2019, has a technology trends document focused solely on AI, and is a whopping 158 pages. I'm going to show you a couple of key figures from this document over the next few minutes. So here's the first interesting graphic. You can see that there's a growth with patent families following scientific publications. The exception to the rule is rule learning, which saw a burst of patent filing around 1999. Now, the reason that this is important, is the number of patent filings usually follows scientific publications. One sign that companies are taking AI seriously are the growth in patents. Here, the ratio of scientific publications to patent families dropped from eight to one in 2010, to three to one in 2016. Now that's really interesting because this suggests a greater interest in the practical uses of AI technologies, and this reinforces that we are in the implementation phase of AI. Machine learning is the dominant AI technique, representing 89% of patent families related to an AI technique. Deep learning demonstrates, by far, the highest recent growth in the field with an impressive 175 percentage average annual growth, between 2013 and 2016. Computer vision and natural language processing, or NLP, and speech processing, feature highly in the number of patents for a given application. And finally, you can see that AI techniques are used in almost every industry vertical. Machine learning, computer vision, NLP, and speech processing, are heavily used in telecoms, transportation, medical sciences, and to manufacturing, amongst a host of others. The darker the green, the higher the number of patent families corresponding to a technology, and a corresponding AI technique. So why bother with AI? Let me give you a couple of concrete applications.

Contents