Artificial Intelligence and the Patent System: Can a New Tool Render a Once Patentable Idea Obvious?

By: William Samore

Introduction: In the summer of 1956, leaders in the field of computer science met at Dartmouth College and founded the field of Artificial Intelligence. Since then, one branch of Artificial Intelligence—Genetic Programming—has progressed to the point where it could drastically change the way that inventors design and create. Genetic programs (described in more detail in section III.B of this paper) operate by mimicking the biological evolutionary process and have a wide variety of applications. Antenna design, for example, is a field where genetic programming could radically change the nature and pace of innovation. The first antennas were built in the late 1800’s by Heinrich Hertz, and an antenna with a specific shape can be designed to emit a desired radiation pattern. As technology progressed, computer programs were designed where an antenna’s characteristics could be inputted to the computer program, and the radiation pattern would be calculated and displayed to the user. Now, computer programs have gone one step further, making it possible to do the reverse: input a desired radiation pattern and have the computer program itself design the antenna. The question that this note asks is, can changes in the tools available to inventors render previously patentable ideas obvious and therefore unpatentable? In other words, should an antenna, which could only have been designed by a human at one point but now can be designed by a computer, be patentable?

Part II introduces the reader to patent law. Part II.A discusses patent law in general, and includes an explanation of the derivation of patent rights. Part II.B then explains the legal concept of obviousness—the most relevant concept to patenting a device designed by a genetic program. Part III discusses relevant technological advances, particularly genetic programming. Next, Part IV argues that when genetic programming becomes widespread in a particular field, advances that could be created by the program should be deemed obvious. To provide a practical application for this argument, Part IV.B sets forth a widespread use test. Part V addresses anticipated contra.

Click here to read the full article