Shortly Governments around the world should pass intellectual property laws granting rights to AI systems, argued two scientists from the University of New South Wales in Australia.
Alexandra George and Toby Walsh, law and AI professors, believe that not acknowledging machines as inventors could have long-term effects on economies and societies.
“If courts and governments decide that inventions made by AI cannot be patented, the impact could be huge,” they wrote in a commentary article released in the nature. “Funders and companies would have less incentive to do useful research with AI inventors if the return on their investment could be limited. Society could be missing out on the development of valuable and life-saving inventions.”
Today’s laws virtually only recognize people as inventors with IP rights protecting them from patent infringement. Attempts to repeal the human-centered laws have failed. Stephen Thaler, a developer who insists AI invented his company’s products, has sued trademark offices in several countries, including the US and UK, to no avail.
George and Walsh support Thaler’s position. “It will not be easy to create tailor-made laws and an international treaty, but not creating them will be even worse. AI is changing the way science is done and inventions are made. We need a fit-for-purpose intellectual property right to ensure it serves the public well,” they wrote.
Dutch police deepfake dead teenager in criminal case
A video clip showing the face of a 13-year-old boy who was shot outside a metro station in the Netherlands and swapped onto a body using AI technology has been released by police.
Sedar Soares died in 2003. Officials failed to solve the case and with Soare’s family’s permission, they deepfaked his image on a child playing football in a field, presumably to refresh someone’s memory. Police have reportedly received dozens of potential leads since then. according to to the guard.
It appears to be the first time AI-generated imagery has been used to solve a crime. “We have not yet checked whether these clues are usable,” said Lillian van Duijvenbode, a spokeswoman for the Rotterdam police.
You can watch the video here.
The AI Task Force advises Congress to fund national computing infrastructure
America’s National Artificial Intelligence Research Resource (NAIRR) called on Congress to launch a “shared research cyber infrastructure” to better provide academics with hardware and data resources for developing machine learning technologies.
The playing field of AI research is uneven. Most modern models are often packed with billions of parameters; Developers need access to many computer chips to train them. For this reason, research in private companies seems to dominate, while academics lag behind at universities.
“We need to ensure that everyone across the country is empowered to conduct cutting-edge research on AI,” the NAIRR wrote in a report. “This growing resource divide has the potential to adversely skew our AI research ecosystem, thereby jeopardizing our nation’s ability to build an AI research community and workforce that reflects the rich diversity of America — and to use AI in ways that are common to all Americans.” serves.”
If AI advances are driven by private companies, it could mean that other types of research areas are left out and underdeveloped. “Expanding and diversifying approaches and applications of AI and opening opportunities for advancement in all scientific fields and disciplines, including in critical areas such as AI verification, testing and evaluation, trustworthy AI, bias reduction and AI safety,” the statement said Task Force argues.
You can read the full report here [PDF].
Meta offers musculoskeletal research technology
Meta AI researchers released Myosuite, a suite of musculoskeletal models and tasks to simulate biomechanical movement of limbs for a wide range of applications.
“The more intelligent an organism is, the more complex its motor behavior can be,” they say said in a blog post. “So an important question to consider is: what enables such complex decision making and the motor controls to execute those decisions? We developed MyoSuite to answer this question.”
Developed in collaboration with researchers from the University of Twente in the Netherlands, Myosuite aims to empower developers working in prosthetics and could help in patient rehabilitation. However, there is another potentially useful use for meta: creating more realistic avatars that can move more naturally in the metaverse.
So far, the models only simulate the movements of arms and hands. Tasks include using machine learning to simulate manipulating the cube or spinning two balls. The use of Myosuite in Meta’s Metaverse is a bit ironic, as touching is not allowed there and hands are restricted to prevent harassment. ®