On an average more than 250,000 patents get filed every year in US alone. Every company worth its salt, especially in hi-tech, is notorious about protecting their intellectual property. They hire teams of lawyers to make sure no one steals or copies their ideas; even if these patents never see light of the day in current or future products of the company. I am confident that the cost-benefit analysis of each patent will be an eye-opening experience for the companies.
Fear-of-being-locked-out is what drives this patent hoarding mentality. What if we introduce a click feature in our product and someone else has already patented the ‘Click’ button? Most of the time, these patents become like cash under the mattress; depreciating asset with no returns. While the patent office can address this issue in a very simple way, AI can help organizations unlock the unrealized potential from dust gathering patents.
In fact, AI models should become an integral part of innovation harvesting framework and identify the ideas that are worth patenting. Multi-variate hypothesis can – 1) Create probabilistic models for predicting product success; 2) Identify an innovation that creates a differentiator for an existing product; 3) New product proposals by collating various ideas from the company’s knowledge and patent repository.
During my tenure with Intel, I managed multiple initiatives on knowledge and ideas harvesting. My experience tells me that structured innovation produces far better results than unstructured innovation. Looking from the outside, that’s what appears to be the key difference between Amazon and Google’s approach to innovation and new ideas.