Risks and opportunities from artificial creativity

Loughran, Roisin (2023) Risks and opportunities from artificial creativity. In: The International Conference on AI and Musical Creativity, 30 August - 1 September 2023, The University of Sussex.

PDF Download (161kB)

Abstract

There has been an explosion of generative algorithms and tools in recent years; applications such as ChatGPT for generating text, Midjourney for generating images or AIVA for generating music have become very popular. As the years pass, we can see an ever-increasing number of generative algorithms and applications being released. Part of their appeal is that such applications are highly accessible to the general public and require little, if any, computational expertise. The ease with which people can generate creative artefacts, such as music, does come with some drawbacks however. There are practical drawbacks to not being able to determine if an artefact was created by a human or not. If we are unable to determine if an artefact is human or computer generated, such methods may cause disruption and difficulties for amateur musicians, artists and creators and for sites or publications who aim to highlight, publish or promote independent artists. Furthermore, many artists within creative domains have shown concern and outright disdain at the notion of generative algorithms being applied to creative tasks. Some of this anger may be attributed to the notion of the humanity of creativity – that machines cannot possibly generate things of genuine beauty, art or creativity. This paper considers a number of practical and ethical questions that are pertinent to the new reality whereby algorithms can be easily used to generate creative artefacts: How can it be determined if an artefact was created by a person or an algorithm? Can an evaluation of the Creativity within an artefact help determine its authorship? Is it necessary to ascertain human versus machine authorship or should the quality of the output be the only issue that matters? What are the repercussions of a person attempting to pass an artefact as human-generated when it was in fact created by an algorithm? While it may not be possible to definitively answer all questions at this time, it is vital to consider these questions and consider the impact and repercussions that such autonomous methods can have on creative authorship.

Item Type:	Conference or Workshop Item (Paper)
Uncontrolled Keywords:	Artificial creativity; Artificial intelligence.
Subjects:	Computer Science
Research Centres:	Regulated Software Research Centre
Depositing User:	Sean McGreal
Date Deposited:	04 Jan 2024 09:30
Last Modified:	04 Jan 2024 09:30
License:	Creative Commons: Attribution-Noncommercial-Share Alike 4.0
URI:	https://eprints.dkit.ie/id/eprint/867

Actions (login required)

View Item

Downloads