AI and Patents: How AI is Creating Super Inventors

By Robert Plotkin and Cynthia Gilbert

This is a special series of articles about the impact of AI on patent protection from Blueshift IP: Software Patent Experts. This series is intended to be easy to understand and to focus on how to tailor AI patents to your business’ goals. This installment focuses on how artificial intelligence is creating super inventors, and the impact of that on patent law.

The key point to take away from this article is that AI can be used by inventors to boost their inventive skills. This point is often overlooked because people so frequently think of AI as something that is replacing human inventors. It’s understandable that the prospect of powerful AI can trigger fear in us about being replaced and made obsolete by machines. This fear is a very old one in our culture, and is reflected in the myth of John Henry, a railroad steel driver who was bested by a steam-powered driver and who died in the process. And there are countless examples of jobs being eliminated by technology.

And AI is no doubt making reducing or eliminating the economic value of many human skills, at least in certain contexts. But our focus on how technology can replace us often causes us to overlook the many other ways in which technology in general, and AI in particular, can be used to augment our skills. Go back a few hundred years and astronomers were bogged down by the need to perform complex calculations by hand in order to plot out the paths of the stars. When improvements in arithmetic, and later in electronic calculators, came about, this didn’t eliminate the need for astronomers. Instead, it freed astronomers to focus on higher-level problems in astronomy and to solve those problems more easily than if they had to spend half their time doing arithmetic by hand. Improvements in calculation helped astronomers become better astronomers.

The same is true today with AI.

For example, a significant part of any inventor’s job is to brainstorm ideas for prototypes and then to build and test those prototypes, learn from the tests, and then go back to the drawing board to make improvements. That process often repeats itself many times before the inventor settles on a mousetrap or piston or drug that works better than anything that came before it.

AI can automate significant parts of this process. For example, if an inventor is trying to design a more aerodynamic airplane wing, many kinds of AI can automatically generate virtual prototypes of potential new airplane wings and even simulate the airflow around them in order to produce at least a rough measurement of how well those wings would perform in the real world. Although the human inventor is still required to provide the initial instructions to the AI, to interpret the results of the AI, and to tweak the AI along the way, the AI can automate significant parts of the process of generating, testing, and evaluating prototypes. Just as in the example of astronomers and arithmetic, AI enables today’s inventors to eliminate a lot of the grunt work involved in inventing and, as a result, to solve the same problem in less time as before, or to solve more challenging problems in the same time as before, because the AI frees up the inventor to focus on higher-level and more complex aspects of the inventive process.

It’s like an inventor with AI has a team of skilled assistants working for him or her. One impact of this is that a single inventor with AI is able to solve problems that would have required a whole team of people in the past. This is one reason we say that AI boosts or augments the inventive skills of inventors.

This has many potential impacts on patent law, but the one that we’ll focus on today is that the concept of inventive skill is baked into a critical aspect of patent law–the requirement that is called nonobviousness in the US and is called inventive step in most other countries. The basic idea is that to be entitled to a patent on a new mousetrap, for example, the inventor has to demonstrate that inventing that mousetrap required more than an “ordinary” level of inventive skill. Although this is very hard to define objectively, the basic idea is that patents shouldn’t be granted for devices or processes that any ordinary inventor, with an ordinary level of skill in his or her field at the time, would have been able to invent. In other words, patents should only be granted for inventions that required more skill than the level of skill that is ordinary in the field.

But if we’re correct that AI boosts the level of inventive skill of inventors who use AI, what does this mean for patent law?  Our view is that if we reach a point where it becomes typical for inventors in a certain field — say mousetrap design — to routinely use AI to assist them in the inventive process, then this would mean that the level of skill of an “ordinary” mousetrap engineer would have been increased, much like if all of them had just earned Ph.D.’s in mousetrap design. And the punchline to all of this is that, if we want the patent system to continue to work in the way in which it is intended, the patent system should take this increase in inventive skill of mousetrap engineers into account when evaluating potential patents for new mousetraps, and only grant such patents if they would not have been obvious for a mousetrap engineer to create with the assistance of AI. The short way to say this is that the widespread use of AI by mousetrap engineers should raise the legal bar for obtaining patents on new mousetraps.

At this point you might be asking yourself, “So what? So what if patent law is oblivious to the ways in which AI is making inventors into super inventors?”  The risk is that patent systems will grant trivial patents, patents on inventions that could have been invented using only ordinary skill. It is a major goal of the patent system not to grant such patents, because of the significant power that patent rights convey. So it is critical that we continue to recalibrate the patent system to the level of skill that exists among different categories of inventors over time, and doing so requires that we take into account the technology–including AI–that inventors use to assist them in the inventive process.

And it’s beyond the scope of this episode to cover all of the logistical challenges that would need to be addressed to adapt the legal system to AI in the way that I’ve described today. That’s a topic for another day!  Our main goal is to convey how AI is augmenting the inventive skill of inventors, rather than making inventors obsolete, and to point out that patent law must take this change in inventive skill into account if patent systems around the world are to stay accurately calibrated to the reality of how inventing is being conducted.

The ways in which AI is making inventors into super-inventors is just one of many impacts that AI is having on patent law and strategy. Stay tuned for the next installment of this special series on AI and patents. If you are using AI to invent in your business and want to talk about obtaining strong, broad, and defensible patents for those inventions, please contact us directly at Blueshiftip.com.

Leave a Reply