Future of the Patent Profession When AI Gains Ground

The following article was written by Alex Paikada, Director Content Marketing at Relecura and was first published on LinkedIn

One of the heavy traffic zones in the emerging technology world is AI (artificial intelligence). Robots and smart machines are progressively taking over many of the highly skilled jobs and hazardous missions. Though nothing in the world is a viable substitute for the human brain, in terms of versatility, reasoning, cognition, and above all, prudence with icing of passion, in certain areas, machines can outperform men, being indefatigable, rugged, and insulated from health hazards. Yet, it is almost certain that the created cannot supersede the creator. However, an angst of sorts is descending on the professional fraternities whether a later day edition of Frankenstein is in the making as more and more missions are being relegated to the obedient, stoic, and no-nonsense phalanx of programmed machines.

Man and machine: the gap is narrowing

That said, AI-propelled machines are increasingly dethroning human labor in many areas of economic activity. In the next few years, the job characteristics and the challenges involved therein will undergo a spectacular transformation. Lately, diagnoses and remote-controlled surgery are being taken over by machines; final decision-making and prescription are left to humane discretion. Certainly, all the hazardous jobs, including rescue operations, jobs in the carcinogenic nuclear facilities, tracing and elimination of terror groups, etc., are areas where those dynamic machines have showed their mettle and grit. Data mining and analysis, among other things, will also be appropriated by smart, autonomous, intelligent beings. In this essentially anthropocentric world, where the end result of every endeavor ought to be the welfare of the species, the usurpation of traditional avenues of engagement to make the daily bread in the sweat of one’s brow by AI, a balancing level ground in between becomes an existential necessity.

AI in IPR data mining and analyses

In the IPR sector, AI, especially ML (machine learning), has already made bold inroads. It is obvious that rummaging through a vast database running to hundreds of millions and very much on the rise second by second, stretching from the late 19th century to the present, is an exasperating Herculean task. In this area, AI has impressively supplemented human intelligence. AI-enhanced semantic searches using neural networks have proved to have high retrieval efficiency and accuracy, and over time, it will scale more heights of perfection. AI and machine learning techniques are designed to assess the behavior of the searcher, compare, and include competitor and market information with patent data.

With AI-powered tools taking over various articulated asks, the following areas are possible to have a conspicuous impact in the long run:

Screening of inventions and carrying out prior-art search will become instantaneous, and manual labor going into those activities, usually running to hours and days, could be saved. Problems of semantics and equivalent words will continue to plague the process for some more time, at least. The limitation of AI is that it has to be guided to receive results that are relevant to the problem at hand. Human discretion will still be needed to deal with the grey spaces.

However, bringing down prior-art search results from tens of thousands of documents to tens of documents with high values of precision and recall is a capacity already made to tremendous use. As of now, the performance and accuracy of the AI-powered tools are brutally dependent on the quality and succinctness of the input data. Describing the inventions in a style easily deciphered by the machines and defining the related areas will be a professional challenge, for which skill tuning and training will be required. Feeding the data to be processed is essentially a professional job that cannot be substituted by machines and diving to the unfathomed recesses and mining out the most pertinent information could become a preserve and prerogative of the smart avatars.

Patent offices are already utilizing AI to screen inventions made by humans. It is going to be the Standard Operating Procedure (SOP) for all practical purposes, subject to subtle fine-tuning and modifications to suit the demands of the day. But human examiners will have to screen the inventions screened by AI. Here, though, the workload will be substantially reduced, and documents to be processed will be distilled down to a dozen or so from thousands.

Inventions made by machines

Inventions made by AI are a new challenge to be dealt with. A new model of patent law in the new era is to be thought of. IP professionals have to design ways to formulate and protect inventions made by machines. This will largely depend on new regulations adopted by various patent offices on handling inventions made by AI. Various patent offices have already initiated steps to introduce standardization and regularization of AI inventions. This is an area that will be warmed up by more trial-and-error phases. The cardinal question raised by the patent offices are: Who will be the statutory owner of an AI invention? Are there any patent eligibility, prior art, and disclosure related considerations unique to AI inventions? Major factors affecting AI inventions are the structure and algorithms of AI systems, the training data, the database of prior knowledge, formulation of a new application to an existing AI system, pre-processing of application-dependent data for the AI system, and post-processing of output from the AI system to the application.

Professional organizations in patent analytics

Patent portfolios are easily accessible to all stakeholders, and data access will cease to be a preserve of the elite. You do not need to spend days or weeks to understand what is in your own patent portfolio or of your competitors. Even when the information dossiers are opened up for public scrutiny and access, the final studied decision-making regarding patentability and competence will continue to be the forte of the IP professionals. An inventor cannot necessarily sit in judgment on their invention with reference to an enormous and puzzling databank, no matter what kind of machines or tools they have access to.

The information dissemination process is going to be faster and easier. The latest tools in this genre from Relecura has revolutionized the area. Tracking or monitoring inventions in your area of work have already been automated, and every development in your area of interest comes to you through reports of a chosen periodicity. This advancement helps the professional abreast and in the know, and naturally, their judgmental skills are honed up without any arduous mobilization.

The protracted and long haul of patent drafting, re-drafting, and emending will, in due course, be taken over by machines. Professionals will still be needed in directing or emphasizing specific parts of the inventions. This level of automation is still a long way away, but the prospects are very much on the horizon. This is, indeed, a core area of the IP profession. But machines cannot entirely replace human intelligence in this activity, given the subtlety and sensitivity of the subject matter intellectual property will become more quantifiable, given the automated means of rating based on smart guiding parameters, and commercialization between a buyer and seller could be on an equal footing as almost nothing is left to conjectures and imagination. The licensee and the licensor will have equal access to information and data-based automated analyses. Hence, decision making would be more rational and based on rigorous ground-truthing. Patent analytics is an area where private organizations are, in recent years, stealing the show. AI-ML-fueled search engines are undergoing swift and economically significant evolution in various parts of the globe. As a result, the traditional niche of the genre of patent agents has already been threatened. Novelty check, prior art search, competitor analysis, graphic description of emerging technologies, tracking technologies, classification of technologies with respect to a control patent, etc., could now be instantly carried out. (For details, visit relecura.com). Nobody is going to turn the clock back; this is a one-way traffic into a bold and beckoning future. In terms of efficiency, accuracy, and precision, more advancements could be expected in the years to come. Given the enormity of the data stack, manual data mining and analyzer are more of a quixotic pipe dream. The end product of this dogged and persistent endeavor will be to automate the standard workflow of Patent Search and Analytics. This situation, as of now, is a long way away, but one could count on it.

Now the seminal question is whether the tyranny of AI will push IP professionals to redundancy and starvation. The stout and the confident answer is NO. WIPO records say that innovators around the world filed 3.17 million patent applications in 2017, representing an eighth consecutive year of growth. Trademark filing activity amounted to 12.39 million, up 26.8%, representing the third consecutive year of double-digit growth. Worldwide industrial design filing activity exceeded 1.24 million, while applications for utility models totaled 1.76 million. Innovation is exploding the world over. The only challenge is that the professionals have to change with the changing times and rehabilitate themselves to a competitive environment presided over by AI. Strenuous manual labor may become out of fashion but taming the machines and goading them to arrive at the right information will be the challenge.

The final smile is whose?

The fundamental truism is ever relevant, though. No machine is a substitute for the human brain. No elephant is autonomous; it needs a mahout to do the job. It requires relevant professional skills, aptitude, and dedication. In addition to the legal, technical, and business skills, an IP professional needs to include the basics of artificial intelligence as well in the arsenal.  Indeed, IP professionals do have a space, for that matter a brave new space, provided they are at home in the changed ecosystem, which calls for a subtle fine-tuning and capacity building. Machines made by man operate and deliver restrained by a set of pre-defined parameters. The machines are at a loss when challenged with an environment not exactly pre-defined. That kind of versatility and flexibility (required in the IP profession) only the human brain is privy to. Also, decision making does not always have to be a brutally rational one. Prudence and temperance, and humanist values are often long-term investments in the business environment. Hence, professionals alone could be the final decision-maker. Machines cannot safely prognosticate your ailment with the symptoms you come out with; a human and humane touch is needed at the final defining moment.  Hence, professional, yours will be the last smile.