A recent body of evidence from a group of 50 global experts has emerged, shedding light on the urgent need to halt the burgeoning growth of technology-driven AI in favour of a more human-centred approach. The experts argue for a shift from innovation at all costs to a model that prioritises the alignment of AI with the needs and values of humanity.
The research, encapsulated in a new book focused on the development of more ‘human-centred’ AI, takes a multidisciplinary approach, with contributions from experts representing over 12 countries and more than 12 fields, such as computer science, education, law, and political science.
Human-centred AI, as it is referred to, aims to promote technological advancement that enriches human existence rather than replaces or devalues it. One prominent researcher, Shannon Vallor from the University of Edinburgh, defines human-centred technology as one that aligns the entire technological ecosystem with the health and well-being of humans. The counter to this, as she points out, is the development of technology designed to replace, compete with, or devalue humans, rather than supporting and empowering them.
The dangers of AI technology as it stands are highlighted in the book, particularly regarding its systemic biases in development. Malwina Anna Wójcik, from the University of Bologna, points out the lack of input from historically marginalized communities in the design and development of AI technologies, which perpetuate existing power dynamics. These biases lead to discrimination and the exacerbation of existing power gaps.
Contributors to the book also delve into the privacy and behavioural implications of current AI use, with particular concern raised around the redefining of privacy in the digital age. Additionally, the impact of AI systems used by social media platforms on users’ range of interests and the potential move toward extremist positions is discussed.
In light of these concerns, the book not only identifies the issues but also proposes actionable solutions. Suggestions include diversity in research, interdisciplinary projects, transparency in company data, and a push for policymakers to extend and apply existing laws to AI rather than creating entirely new legislation.
Policymakers are called upon to shift focus from an open-ended ethical discussion to actionable policy-making, with suggestions to instil confidence in regulating AI and to involve a wide range of perspectives, including marginalized communities and end users in governance mechanisms.
Overall, the book serves as a rallying call for a paradigm shift towards human-centred AI, advocating for a more conscientious and interdisciplinary approach in both policy and practice concerning AI development. The detailed recommendations and insights presented lay the groundwork for a more equitable and responsible future in the era of artificial intelligence.
Reference: “Human-Centred AI – A Multidisciplinary Perspective for Policy-Makers, Auditors, and Users”, 21 March 2024. DOI: 10.1201/9781003320791
+ There are no comments
Add yours