A federal court recently ruled that an AI can’t be listed as an inventor on a US patent. The judge upheld a decision that the machine does not qualify as an inventor because it is not a person. But the case also plays into the murky issue of whether or not computers can be creative.
Creator or Created?
The recent court case proves that AI may not be getting all the credit for what it can do. Computer scientist Stephen Thaler built an AI “creativity machine” called DABUS, which stands for Device for the Autonomous Bootstrapping of Unified Sentience. Thaler’s company, Imagination Engines Inc., filed a patent application in 2019 that listed DABUS as an inventor of a “neural flame” device containing a flashing light-emitting element and a beverage container created using fractal geometry. DABUS can invent things and deserves the patents, Thaler claims. But Judge Leonie Brinkema ruled in the case that, under US law, only a human can be an inventor. “As technology evolves, there may come a time when artificial intelligence reaches a level of sophistication such that it might satisfy accepted meanings of inventorship,” Brinkema wrote in her decision. “With AWS DeepComposer, developers are able to use the console to pick a music genre and provide their own input using the keyboard to generate entirely new music using sampling models,” Miller said.
AI Could Boost Human Brain Power
The ruling by Judge Brinkema against AI inventors is rooted in the letter of patent law. But the courts have not tackled the more complicated task of determining whether an AI invention is truly unique or a reproduction, James Kaplan, CEO of MeetKai, which makes an AI voice assistant, told Lifewire in an email interview. “I would argue that as long as a patent reviewer grants the patent with a name-blind inventor, then it passes that smell test,” he added. Kaplan said that AI would never be the “sole” inventor on a patent; instead, humans will always be in the loop. People will sketch out a problem, and the AI will be in charge of suggesting and helping to fill in the blanks. “We have already seen the early signs of this in programming, where new models available today can generate code given plain text descriptions of the desired outcome,” he said. Joseph Nwankpa, a professor of information systems and analytics at Miami University in Ohio, agrees that AI has not advanced to the level where it can be considered an inventor. “It is still unclear how to define computer autonomy during the inventive process,” he told Lifewire in an email interview. “However, as AI develops, it will inevitably reach that high intelligence threshold and attain the capability to make an invention independently.” AI is likely to drive future innovations, Miller said. For example, to aid in drug discovery, scientists recently have used AI to create new drug-like small molecules targeting novel viral proteins. AI could even create original artworks based on the work of human painters, David De Cremer, the director of the Centre on AI Technology for Humankind at the National University of Singapore, told Lifewire in an email interview. “AI is very well able to make new paintings that are entirely in the style of these masters of the past, and people are willing to pay big bucks for it,” he said. “Until they find out that AI made it, then it suddenly loses its value in people’s eyes.”
