Dostęp otwarty
FACING THE ALIGNMENT PROBLEM THROUGH NIMBLE CHANGE-MANAGEMENT. LIMITATIONS OF LEGAL ENVIRONMENT IN COPING WITH DIGITAL AGENTS
Piotr BOŁTUĆ1
Język publikacji: angielski
Artykuł w czasopiśmie
Transformacje Nr 4 (119) 2023, Data publikacji: 29 grudnia 2023r.
Słowa kluczowe: The Alignment Problem; Human AI Gemeínschaft; Justice Birss; Justice Beach; DABUS
Streszczenie The Alignment Problem, asking how to align the goals of AGI with human in a long time-frame, assumes AGIs having ‘their’ goals. I view the alignment as a possibility, only when the human-AGI arrangement comes from the community, and it is not based primarily on controls, or laws. The sole way to create community of purpose is by forming Gemeínschaft, with AGIs as members. In such community, humans should not be ashamed of their inferior intelligence, or even strategic skills, while Advanced AIs (not just robots, but also advanced cognitive engines) should not be discriminated against, as ‘aliens’ The case of DABUS, which was quite obviously an inventor of certain patentable artifacts, has been adjudicated by many of the most inluential courts on the globe. All the courts (except for RSA), ruled against the possibility of DABUS being an inventor, since it is not a human person. This demonstrates how far from Human-AI Gemeínschaft our legislators, as well as international patent treaties, operate. We can do better on the political, legislative and legal side. Fortunately, Justice Birss (UK), and Justice Beach (AU) provide examples how to challenge philosophical status quo withing legal theory and legislative practice – the former, by treating historical roots of patent law with due respect; the latter, by taking technological progress quite seriously. The other way which seems to work in RSA, follows Occham, having the courts leave aside philosophical anthropologies.
0000-0002-7210-615X
ORCID: pboltu@sgh.edu.pl
E-mail: