It’s possible, as some fear (and some popular movies have predicted), that computers will ultimately become self-aware and turn on their human creators. But if so, they must first establish that they’re legally capable of creation. Put differently, perhaps we should prioritize the new CGI (computer-generated inventions) over the old (computer-generated imagery).

I’ve previously examined the general landscape of inventions by machines and the cases for and against granting them patents.

As we saw, Missouri-based inventor Stephen Thaler has pioneered the development of CGI, in particular with “DABUS,” the Device for the Autonomous Bootstrapping of Unified Sentience, which Thaler characterizes as “a swarm of many disconnected neural nets.”

DABUS essentially scours the world for potential inventions
— i.e., innovative ideas that no one has yet conceived of — and then critiques
and double-checks the viability of those inventions. I recently spoke with
Reuven Mouallem, an Israeli patent attorney and US patent agent, who has filed
patent applications on behalf of DABUS in the US and Israel as part of the
Artificial Inventor Project — a team of international patent attorneys seeking intellectual
property rights for the autonomous output of artificial intelligence (AI). Dr.
Mouallem told me this critiquing function “lends DABUS a quality most akin to
sentience, a chaining of conceptual hives resembling the formation of a series
of memories.”

A few years back, Thaler filed several patent applications
on behalf of DABUS in the US, the UK, Europe, Taiwan, Vietnam, and elsewhere,
claiming inventions ranging from a food container graspable by robots to an
emergency warning light. But patent offices across the globe have thus far
rejected all of DABUS’s applications on the grounds that the applicable laws
require an inventor be a natural person.

Thaler and DABUS have appealed these orders, but in September, the United Kingdom’s High Court of Justice affirmed the rejections, finding that “in this case DABUS is not the applicant: Dr Thaler is,” and UK law requires human inventorship (emphasis in original).

The High Court did, however, note that it “in no way
regard[s] the argument that the owner/controller of an artificially intelligent
machine is the ‘actual deviser of the invention’ as an improper one.” Yet this
observation, presented in a postscript to the decision, is unlikely to satisfy
Thaler, who seeks inventorship for DABUS, not credit for “devising” its
inventions.

In addition, a lower tribunal found that “neither the Patents Act 1977 nor the European Patent Convention explicitly prohibits protection for autonomous machine inventions” and that the “natural person” requirement was “intended to prevent company inventorship.”

As Mouallem explains, “These statutes were designed to
prevent companies from claiming inventions without crediting the individuals
actually responsible for them. However, crediting an AI with an invention is
perfectly consistent with these goals.” In other words, there’s room for
British lawmakers to revise patent legislations to allow for CGI.

By contrast, in China, newly issued draft patent examination guidelines specifically forbid awarding inventor status to AI. These draft rules, however, remain subject to public comments for another month, and DABUS and its supporters are expected to challenge this provision.

Meanwhile, last month, the US Patent and Trademark Office released a report titled “Public views on artificial intelligence and intellectual property policy,” collating some 200 comments from the interested members of the public. “The majority of comments,” per the report, “suggested that current AI could neither invent nor author without human intervention” and “expressed a general sense that the existing U.S. intellectual property laws are calibrated correctly to address the evolution of AI.”

So DABUS and its advocates still seem to have their work cut out for them. It may yet be a while before Skynet becomes self-aware.

The post More adventures in CGI (computer-generated inventions) appeared first on American Enterprise Institute – AEI.