Among the many issues under discussion regarding the possible impacts of artificial intelligence (AI) on humanity are those related to the production and use of knowledge. This year's Nobel Prize awards focused on advances that are directly related to AI and computer science. Apparently, nothing will be done without the use of tools that, even in the near future, we do not even know what they will be.
I cannot recall an innovation that has had such rapid dissemination and impact as ChatGPT, which, launched at the end of 2022, immediately became a global reference and triggered the rise of other tools based on Large Language Models. And, experts say, we are only at the beginning of a series of transformations whose unprecedented nature makes predictions even more difficult.
Creating things, as we know, depends on new knowledge, which, as we also know, is largely based on science and technological development. Sometimes it starts in laboratories, sometimes in inventive practice. To date, the creations that have had the greatest impact on humanity have come from scientific research and the entrepreneurial and visionary capacity of a few.
With AI, some immediate issues to be faced in the world of science and technology are authorship and reproducibility of creations, two foundations of science and technology governance.
Who should be credited with authorship and ownership of something created by a generative AI? How can we guarantee the reproducibility of a model that, with each use, learns and changes and is therefore no longer the same?
Today, the answer, as we will see below, has been pragmatism, since autonomous creations and inventions, although not yet circulating among us, are already occurring in laboratories, as is the case with the creation of new technologies by AI. antibiotics to combat resistant bacteria. As for the future (which seems not so distant), it is impossible to predict the answer, especially because we do not know what could be created, invented and how far it will be from the original event.
On the issue of authorship, as discussed by the authors of the article “A Shift in the World of Science”, published in the New York Times (NYT) on October 13, scientists may produce the tools that will later make advances in knowledge, advances of which they themselves will no longer be the authors or inventors. In the world discussed in that article, the inventors of algorithms and tools capable of learning and creating will no longer do the revolutionary work and may even lose the attribution of causality over what will be done with the tool they developed. Perhaps they will be the precursors of this work, like the character Eldon Tyrrel, of the Tyrrel Corporation, in the now classic science fiction film Blade Runner.
In the NYT article, the authors (who are not machines, I believe) ask themselves: “Who holds the discovery? Where does the machine end and the human begin?” They emphasize that, today, science can increasingly be considered the result of collective efforts, and that the AI tool used by the researchers who won this year’s Nobel Prize in Chemistry was trained using a database that brought together the work of more than 30 biologists. Also in the case mentioned above, the development of antibiotics by AI, it was necessary to gather a large amount of data (and researchers). This, in fact, is one of the great challenges for the adoption of AI in scientific and technological research: the need for large databases of research results and advances that can be combined in the training of the tools.
In the field of intellectual property (IP), the issue of how to treat AI-based inventions is perhaps the most important issue in patent offices around the world. IP is a highly institutionalized issue, codified and governed by national standards and some international standards endorsed by national standards. National patent offices (such as the United States Patent Office – USPTO, the European Office and the National Institute of Intellectual Property – INPI in Brazil), which also operate other forms of IP protection, have clear rules on what can and cannot be characterized as an invention eligible for intangible property protection.
All consider that a patentable creation must be the result of human inventiveness (in addition to originality and industrial applicability), and none (to date) considers attributing ownership to a non-human entity. The text presented by Guerra et al. (2023)[1] shows that none of the major national offices (United States, European Union and China), besides Brazil, had, until the end of 2023, specific rules for creations resulting from AI.
But, as this is a subject in constant and rapid evolution, in February 2024, the USPTO published guidelines for patenting AI-assisted inventions. The main discussion of these guidelines concerns the need for there to be a natural person for ownership to be attributed to an invention. The text states:[2]
“Although AI systems and other non-natural persons cannot be listed as inventors in patent applications or patents, the use of an AI system by a natural person does not prevent that natural person from qualifying as an inventor (or joint inventors) if he or she has contributed significantly to the claimed invention (…) Patent applications and patents for AI-assisted inventions must name the natural person(s) who contributed significantly to the invention as inventor(s) or joint inventors.”
“(…) Likewise, the Federal Circuit has made it clear that conception is the fundamental point of inventiveness (…) Since conception is an act carried out in the mind, until now it has been understood as something carried out only by natural persons”.
So far, as the text says, a generative AI that creates something new through its own learning and that is distant from its creator has no legal provision different from the current one, neither with the USPTO nor with other national offices. [3]
The US guidelines document, like those of other countries, in the face of such uncertainty, adopts a pragmatic stance:
“The USPTO will continue to presume that the inventor(s) named in an application are the actual inventors. And applicants will remain responsible for fulfilling their existing obligations to the USPTO. Only in rare cases, in which an examiner determines (…) that one or more of the named inventors may not have invented the claimed subject matter, would questions of inventiveness be raised during review. From an examiner’s perspective, it will not matter whether the AI or other advanced computer system performed actions that could rise to the level of inventiveness. What matters, according to the guidelines, is whether the actions of at least one human can be shown to be sufficient to reach the level of inventiveness (…).”
In other words, that's it for today.
In the words of a former US Secretary of Defense, famous for an action that was not exactly an example of success (the invasion of Iraq in the early 2000s), “there are things we know that we do not know, and others we do not know that we do not know”.
The future of AI is in both sentences, but it fits more in the second one. Creating things that we don't even know can exist is, perhaps, the biggest unknown that we have to face with the emergence of AI in science and technology.
That's what we have for tomorrow.
Observation: I would like to thank Professor Anderson de Rezende Rocha from the Institute of Computing (IC) at Unicamp, who suggested important adjustments and topics for the text. Any errors and inaccuracies that persist are my sole responsibility.
This text does not necessarily reflect the opinion of Unicamp.
[1] See pages 2123 to 2132 of: https://www.altecasociacion.org/_files/ugd/9d974b_5fa00e9cdfd64130ae3042efe321406f.pdf
[2] Translation made on ChatGPT and reviewed by this author.
[3] Another guidelines document is that of the Singapore IP Office, in which there is discussion based on concrete situations that point to potential changes in the regulatory and jurisprudential framework, although the document itself is not binding nor does it replace the legislation in force in that country.
