Asma Raza

Who is Winning the Artificial Intelligence Race?

“The U.S. Patent and Trademark Office in 2019 conceded 14,838 licenses that referenced AI or ML, of which 1,275 explicitly referenced AI or ML in their titles or modified works. That is generally twofold the issuance in 2018.”

Much has been expounded on how man-made consciousness (AI) and AI (ML) are going to change the worldwide profitability, working examples and ways of life and make colossal riches. Gartner ventures that by 2021, AI growth will make $2.9 trillion of business esteem and $6.2 billion hours of laborer efficiency all inclusive. McKinsey figures AI conceivably could convey extra monetary yield of around $13 trillion by 2030, boosting worldwide GDP by about 1.2 percent a year. Organizations around the world are on the whole hustling to embrace and improve AI and ML advances. Without a doubt, by any record, much advancement has been made and the appropriation and development rates are enlivening. In any case, who is winning or driving in the race? A speedy survey of U.S. patent information may give a look into the condition of the race.

Delving Into the Data

The U.S. Patent and Trademark Office (USPTO) in 2019 allowed 14,838 licenses that referenced AI or ML, of which 1,275 explicitly referenced AI or ML in their titles or modified works. That is generally twofold the issuance in 2018, where 8,227 conceded licenses referenced AI or ML, and 515 explicitly referenced AI or ML in their titles or digests.

The U.S. Simulated intelligence/ML licenses allowed in 2019 spread a wide scope of regions, from selection and utilization of AI/ML advances to life science, designing, registering, web based business, to business/account to development in machine preparing and neural system advances themselves. As anyone might expect, order 706, information handling: man-made consciousness, has the most elevated number of licenses allowed that explicitly referenced AI/ML in either the title or unique, at 151. Instances of class 706 licenses are U.S. Patent No. 10198399, Cryptographically Secure Machine Learning, and U.S. Patent No. 10198698, Machine Learning Auto Completion of Field. Different classes with in excess of 50 licenses allowed that explicitly referenced AI/ML in either the title or theoretical are:

  • Class 713 Electrical Computer and Digital Data Processing and System Support: 107
  • Class 209 Classifying, Separating and Assorting Solids: 93
  • Class 455 Telecommunications: 85
  • Class 701 Data Processing: Vehicles, Navigation and Relative Location: 84
  • Class 705 Data Processing: Financial, Business Practice or Cost/Price Determination: 64
  • Class 340 Communications, Electrical: 58

The Art Units intended to deal with these applications are:

  • Craftsmanship Unit 2129 for Class 706
  • Craftsmanship Unit 2131 for Class 713
  • Craftsmanship Units 3642 and 2649 for Class 455
  • Craftsmanship Unit 3655 for Class 701
  • Craftsmanship Units 3685, 3689 and 3693 for Class 705
  • Craftsmanship Units 2481, 2684 and 2686 for Class 340

The normal time it took from documenting to issuance was 850 days. The patent that set aside the longest effort to issue was U.S. Patent No. 10410308, System, technique, and gadget for individual therapeutic consideration, insightful investigation, and finding. It took 4,530 days. The quickest issuance was U.S. Patent No. 10470510, Systems and Methods for Full Body Measurements Extraction Using Multiple Deep Learning Networks for Body Feature Measurements. It took just 94 days.

Who’s Leading the Pack?

As anyone might expect, Big Techs are the top beneficiaries of these licenses, with IBM standing out; the organization got 81 of these licenses. Following IBM are Microsoft with 56, Amazon with 51, Cisco and Facebook each with 30, and Google with 26. Apple came in at a removed 6th spot, having gotten just 10 of these licenses. Maybe that clarifies to some degree Apple’s ongoing acquisition of Seattle’s Xnor.ai for $200 million, after its prominent acquisition of another Seattle startup, Turi, in 2016.

In any case, Big Tech organizations are by all account not the only significant beneficiaries of these licenses. Among the non-Big Tech firms, Capital One got 50, Fanuc Corp got 36, Accenture got 21, and Bank of America got 15 AI/ML licenses. Huge numbers of these are reception licenses, as one Capital’s “Using AI with self-bolster activities to decide bolster line positions for help calls” U.S. Patent No. 10263862. Others are innovation driven licenses, similar to “Frameworks and strategies for quickening model preparing in AI” U.S. Patent No. 10332035, and Systems and strategies for furnishing robotized characteristic language exchange with clients U.S. Patent No. 10322505.

For non-U.S. organizations, European monster Siemen stands out, accepting 18 such licenses, trailed by Korean’s LG with 11 and Samsung with 9. Among Japanese organizations, NEC was the pioneer, accepting 6 AI/ML licenses, trailed by Sony with 7 and Toyota with 2.

Eye on China

A great deal has been expounded on China’s attention on AI and how the AI/ML filings in China by Chinese organizations have fundamentally expanded. Maybe in view of the time it takes from recording to issuance, we are yet to see the Chinese push in gave U.S. licenses. Of the Chinese Big Tech names, Baidu stands out, getting 19 of these licenses. Huawei got just 2, and Tencent got just 1. Be that as it may, one marvels if the image may be essentially unique in 2020 and past.

Leave a Reply

Your email address will not be published. Required fields are marked *