US-based artist and engineer Alexander Reben has invented an invention which invents.

His latest project, named All Prior Art, is an algorithm which produces ‘new’ descriptions of theoretical inventions by rearranging and combining extracts from the US patent database.

A glance over the most recently published ideas exposes a monkeys-on-a-typewriter approach but beneath the computer generated technobabble is a serious goal: “most inventions generated will be nonsensical” Reben admits, but “[t]he concept is to democratize ideas, provide an impetus for change in the patent system, and to preempt patent trolls.”

Reben’s objective is to publish every conceivable remaining innovation before they can be patented. Part of the reason is to prevent patent trolls (dishonest companies and individuals who bring often-frivolous infringement claims for the purpose of extorting settlement fees) from registering vague, ambiguous or plain-obvious patents to the detriment of genuine innovation. On a more fundamental level, the project also aims to undermine the monopolisation principle which underpins the modern patent system.

By way of a refresher, patents are time-limited monopolies granted over an innovation in exchange for the inventor revealing their innovation to the world. In straightforward terms, the inventor enjoys the exclusive right to sell the idea and the public benefits from the idea being disclosed in full, allowing others to build upon it. A disclosed innovation forms part of the ‘state of the art’. The ‘prior art’, the project’s namesake, refers to the state of the art prior to the date a patent application is submitted. A patent application for an idea which already exists in the prior art would (in principle) be refused for lack of innovation.

Herein lies where All Prior Art’s pages of randomised jargon might not be so harmless. Behind the endless word-scramble is a risk that the algorithm will churn out a specification of someone’s work-in-progress. If this happens then the innovation will not be patentable as the idea already exists in the state of the art.

Without patent protection, an inventor releasing their product onto the open market would enjoy a limited period of exclusivity before even the most complex system were reverse-engineered and copied. Without a patent, the inventor would have little recourse to others exploiting their idea and copying parties would undeservedly reap the reward of the inventor’s costly research and development.

Reben’s activism against the American patent system also has global implications. Although his concerns regarding the American patent system may not apply to other jurisdictions, the engineer’s computer-generated innovations have the potential to undermine attempts to protect legitimate innovation world-wide as the state of the art is established at a global level. An idea-by-algorithm published on All Prior Art’s website has real potential to undermine a genuine patent application in the UK (or any other territory globally) which is based on similar wording.

Is there a cause for concern then? Theoretically, given sufficient computational power and time, the All Prior Art algorithm could produce every possible innovation conceivable, effectively inventing everything that could possibly be invented. The time needed for such an achievement, however? Infinite. On a practical level, it is very unlikely that your innovation will be disclosed by the project in the foreseeable future. If the reader has any second thoughts, a glance through some of All Prior Art’s most recent publications should give considerable peace of mind.