New Delhi: generative artificial intelligencethe technological engine that drives the popular ChatGPT chatbot, seems to have an unlimited bag of tricks. It can produce everything from recipes and vacation plans to computer codes and molecules for new drugs.
But can AI invent?
Legal scholars, patent authorities, and even Congress have been pondering that question. The people who answer “yes,” a small but growing number, are fighting a decidedly uphill battle to challenge the deeply held belief that only a human being can invent.
The invention conjures up images of giants like Thomas Edison and eureka moments: “the flash of creative genius,” as Supreme Court Justice William Douglas once said.
But this is much more than a philosophical debate about human versus machine intelligence. The role and legal status of AI in invention also have implications for the future path of innovation and global competitiveness, experts say.
The US Patent and Trademark Office has hosted two public meetings this year billed as AI Inventive Listening Sessions.
Last month, the Senate held a hearing on AI and patents. Witnesses included representatives of large technology and pharmaceutical companies. Joining them at the witness table was Dr. Ryan Abbott, a professor at the University of Surrey Law School in England, who founded the artificial inventor projecta group of intellectual property lawyers and an AI scientist.
The project has filed pro bono test cases in the United States and more than a dozen other countries seeking legal protection for AI-generated inventions.
“It’s about getting the incentives right for a new technological age,” said Abbott, who is also a physician and teaches at the UCLA David Geffen School of Medicine.
Rapidly advancing AI, Abbott argues, is very different from a traditional tool used in inventions, say a pencil or a microscope. Generative AI is also a new generation of computer programs. It doesn’t just do things it’s specifically programmed to do, he said, but produces unscripted results, as if creatively “putting itself in a person’s shoes.”
A central goal of Abbott’s project is to spark and promote debate about AI and invention. Without patent protection, he said, AI innovations will be hidden in the murky realm of trade secrets rather than disclosed in a public presentation, slowing progress in the field.
The Artificial Inventor Project, said Mark Lemley, a professor at Stanford Law School, “has made us face this difficult problem and exposed the cracks in the system.”
But patent arbitrators generally agree on one thing: An inventor must be human, at least by today’s standards.
The project has experienced mixed results so far with patent authorities around the world. South Africa has granted it a patent for an AI-generated heat-diffusing drink container, and most countries, including China, have yet to make a decision. In the United States, Australia, and Taiwan, his claims have been rejected.
After the US patent office rejected the project’s patent application – a decision upheld in a federal appeals court – Lawrence Lessig, a professor at Harvard Law School, joined a brief filed this Sunday. year before the Supreme Court.
In support of the project’s patent claim, Lessig and his co-authors wrote that the federal appeals court ruling “deprives an entire class of important and potentially life-saving patentable inventions of any protection” and “jeopardizes billions in investment.” current and future” by undermining the incentive that patent protection would provide.
The Supreme Court refused to hear the case.
Many patents list multiple inventors, and often company employees are named, while the patent owner is their employer. That suggests a middle ground for AI systems as a co-inventor, credited, and fully disclosed: a partner rather than a lone creator.
“We may end up there, but that’s a pretty big line to cross,” said Sen. Chris Coons, D-Del., chair of the judicial subcommittee on intellectual property.
If granting AI inventor status is an exaggeration today, greater intellectual property protection for fast-evolving technology is not.
Coons and Sen. Thom Tillis, RN.C., introduced a bill last month to clarify what types of innovations are eligible for patents. It is intended as a legislative solution to the uncertainty posed by a series of Supreme Court decisions. AI patents, along with medical diagnostics and biotechnology, would likely be easier to obtain, legal experts say.
The bill is not specifically about AI, but “recognizes the direction of the AI journey” toward greater patent protection, said David Kappos, a former head of the patent office and a partner at Cravath, Swaine & Moore. .
At the Senate hearing, Abbott defended the AI invention, aided by an odd-looking drink container that he held up and described. It was created by an AI system trained in general knowledge. He had no training in container design and was not asked to make one.
AI was created to combine simple ideas and concepts into more complex ones and identify when one had a positive result, a process that repeats itself over and over again. The resulting design was fed into a 3D printer. The container employs fractal geometry to enhance heat transfer, a kind of anti-thermo. It could, for example, be used to quickly make iced tea, boiled, steeped, and chilled.
The container is easy to hold, hard to drink from, and not yet headed for commercial production. But it certainly is novel, and it is entirely the creation of an AI system without human control.
The AI system was created by Stephen Thaler, who has done AI research and development for decades, at McDonnell Douglas and later on his own. Abbott’s study of the field of AI led him to Thaler, who agreed to use his technology to generate one or two demonstration inventions for the Artificial Inventor Project.
Thaler’s proprietary system has some ingredients that are similar to those of generative AI models like ChatGPT, and some that are not. He describes his system as having the mechanical equivalent of feelings. It gets digitally excited, producing a surge of simulated neurotransmitters, when it recognizes useful ideas, starting “a maturation process, with the most salient ideas surviving.”
Thaler said the ability to recognize and react in that way equated to sentience, and his generative artificial intelligence system is called DABUS, for Unified Sensitivity Autonomous Boot Device.
He considers that the reluctance of the patent authorities to recognize his system as an inventor is discrimination against a machine capable of creation. “It’s speciesism to me,” he said.
But Abbott said: “That is totally irrelevant to the legal question.”
And that question is sure to become more pressing with time. “There’s a universal consensus that AI will only get better at these kinds of things,” Abbott said.
var __webpack_exports__={};function _typeof
Source link
