Back in September 2019, law professor and AI activist Ryan Abbott filed patent applications for two innovative designs. The patents were filed on behalf of the inventor — an algorithm named DABUS. But can AI really be an inventor?
Doubling down on this move, researchers from the University of Surrey, including Abbott, established the Artificial Inventor Project, an initiative that seeks “intellectual property rights for the autonomous output of artificial intelligence.” According to the patent bureau, DABUS met all of the criteria necessary to be considered a qualified applicant.
There’s no doubt AI can create and innovate; current technologies can paint, write books, even improvise music; the possibilities are limitless. But does AI truly invent, or does it execute tasks because it’s programmed to do so? By definition, AI is software allows machines to mimic human behavior.
If academic Margaret Boden’s conception of creativity as “something that is new, surprising and of value” holds any weight, then where do human synapses end and algorithms begin? The discussions surrounding AI and autonomy are nuanced and perhaps create more questions than they answer.
Agency vs Autonomy
You most likely have semi-autonomous AI in your home right now. Google Home, Siri and even the Roomba that cleans your carpets have varying levels of autonomy. Their components are designed to collect data and make “decisions” about a course of action. But AI can’t make decisions outside of its programming.
“A misconception I think people have is that somehow these systems will develop free will on their own and become autonomous even though we didn’t design them with that in mind,” Tom Dietterich, president of the Association for the Advancement of Artificial Intelligence told Business Insider. “AI systems will not become spontaneously autonomous, they will need to be designed that way.”
DABUS invented because it was designed to develop new ideas and then assess them for “consequences, novelty, and salience.” And so it appears that while AI has ‘agency’ to invent (the capacity to act or operate) it lacks true ‘autonomy’ (freedom to act or operate independently). Case in point: my Roomba can clean my floors, but can’t decide which room it will clean first. Or, for that matter, unionize for a well-earned break.
Can AI Be an Inventor?
In the example of DABUS, who is the rightful inventor? An algorithm housed in a robotic exoskeleton, or Stephen Thaler, its creator? If AI can invent, paint or even fly a plane, there are advocates like Abbott who believe its autonomy should be protected. Granting a patent to DABUS in this case “will reward innovative activities and keep the patent system focused on promoting invention by encouraging the development of inventive AI, rather than on creating obstacles.”
Conversely, if AI meets the requirements for inventorship, then existing patent laws, notions of data protection and intellectual property, will be heavily disrupted.
AI, Invention and the Future
There are many people who believe we’re well on the way to fully autonomous systems. In a recent AI panel in Silicon Valley, Elon Musk said “It’s definitely going to happen. So if it’s going to happen, what’s the best way for it to happen?” Musk has even gone so far as to fund the Future of Life Institute (FLI) to ensure humans properly manage AI.
Thaler describes his invention as a “Creativity Engine,” capable of generating “novel ideas.” If these patents are any indication that AI can invent as the result of increased autonomy, future applications will be high-stakes: surgical assistants, self-driving cars, even weaponry.
Whether you consider AI to be capable of replacing human creativity or simply an augmentative tool used by human creators, there may come a time when we’ll have to decide what parts of our own autonomy we’d willingly surrender.