AI in IP: What's Next, Part One

  • Added
  • Author:
IFI attended and sponsored the 2018 IP Service World (IPSW) in Munich where a lot of interesting products and ideas involving artificial intelligence (AI) were discussed. This sparked curiosity around the topic, and we’ve been wondering where the future might take us in regard to its use in intellectual property. Toward this end, please join us at a roundtable event at the 2019 IPSW on the morning of Tuesday, November 26.
 
 

AI as an Assistant

 
Many products are now incorporating machine learning to manage the ever-growing volume of IP records. It’s a necessity now that there are 14 million active patents and 50 million active trademarks worldwide, according to the World Intellectual Property Organization (WIPO). Even highly skilled researchers face nearly impossible odds to find relevant documents without the help of basic AI tools.
 

AI as an Interpreter

 
Dealing with records in multiple languages has always been a challenge. This is why IFI recently made the switch to Google Translate for the most accurate translations available on the market today. Google’s neural machine translation leverages deep learning to evaluate not just words and phrases but entire sentences and paragraphs while considering its context. This approach results in a 60% improvement in benchmarked results over older methods.
 

AI as an Inventor

 
In addition to search, AI is calling into question who or what can be considered an inventor on a patent application. A recent patent application to the EPO, US, and UK lists an AI inventor called Dabus (device for the autonomous bootstrapping of unified sentience) as the creator of a plastic food container and blinking light. In Europe and the UK, the current law states an invention must be made by “natural persons”. In the US, the inventor must be an individual. This development has opened a floodgate at the respective patent offices about how to handle AI.
 

AI as Intelligence

 
While current uses of AI mostly involve machine learning, future developments could make use of deep reasoning functionality. While still in its infancy, this is the type of technology that futuristic movies love to speculate about. Right now, AI is only able to take in large amounts of data and make inferences about other data sets matching the original data. Deep reasoning will represent a jump towards applying learnings to different subjects. IBM is writing extensively on what the future may bring and gave a thought-provoking lecture at IPSW last year. It will be interesting to see what this year’s speakers have to say on the topic.
 

What Do You Think?

 
As AI technology continues to evolve, this interactive event will gather our collective knowledge on what new advancements have taken place and where the future may lead us. Shortly after IP Service World, IFI will publish a follow up blog post describing the conversations which took place at the roundtables with thoughts and ideas from participants.

Read part two with comments from IP Service World Attendees
Edited