New IBM patent uses IoT devices, eye-tracking, mics, and cameras to help make you a better writer

IBM was awarded a new patent that uses cognitive systems and invasive monitoring to improve writing composition via automated behavioral and content recommendations.

IBM

Google has a patent on Contextually Relevant Knowledge Panels in an Application Environment, a method currently used in Google Docs. Google analyzes written text and then displays relevant content that can be added to the document.

In March 2021, IBM was awarded a patent that makes Google’s patent look rudimentary.

The newly awarded IBM patent, Automated Document Authoring Assistant Through Cognitive Computing (2.7MB PDF), seems to pick up where Google left off. Google’s patent suggests related content, but IBM’s cognitive computing system checks the validity of the content and makes suggestions based on user behavior and big data analysis.

One of the visual examples provided in IBM’s patent shows how their system might notify the user that they’re using the wrong image.

IBM Document Authoring Tool Image Correction
An example of IBM’s AI determining the author is using or describing the wrong image

IBM’s patent goes well beyond correcting images. It covers every aspect of the authorship journey.

The computer-implemented document authoring may monitor the narrative evolution by controlling authoring styles, narrative mile stones and subjects that should be covered. This monitoring may be done in real-time through artificial intelligence (AI) techniques, for example, including Machine Learning classifiers, Natural Language Processing (NLP), and Computer Vision (CV), to infer content semantics from authors’ input text and visual data.

Another visual example in the patent portrayed the potential of IBM’s authoring assistant. It showed how it could use its comprehension of a document to automatically analyze tens of thousands of similar papers in the background, check for inconsistencies, and then make recommendations.

IBM patent example of making document suggestions
An example of IBM AI analyzing more than 20,000 paper to make a recommendation

In addition to analyzing what the user writes in a document, the patent also describes monitoring the author with cameras, eye-tracking, mics, and internet of things (IoT) devices to better understand their behaviors and preferences. It then uses the data to make content recommendations and suggest the best time for them to write.

The patent as a whole presents a futuristic writing experience where IBM’s AI is capable of performing superhuman tasks to help people produce the best documents possible. However, you’ll have to get comfortable with it watching your every move.

Related News

Jon is the founder and Managing Editor of Coywolf. He has over 25 years of experience in web development, SaaS, internet strategy, digital marketing, and entrepreneurship. Follow @henshaw