A team of experts is advocating for prioritization of human-centered technology over the indiscriminate development of artificial intelligence (AI). Researchers emphasize the importance of creating technology that addresses human needs and improves the human experience, rather than solely focusing on advancing computer capabilities.
The new book “Human-Centered AI” features insights from 50 experts across 12 countries and diverse fields such as computer science, law, and sociology. It advocates for transitioning AI from a technology-centric approach to one that prioritizes human well-being. This shift aims to ensure that AI advancements complement rather than replace or undermine human labor.
Shannon Vallor, a prominent expert at the University of Edinburgh, stresses the importance of human-centered AI, which prioritizes supporting and empowering humans over showcasing technological prowess. She critiques the development of generative AI, arguing that it often serves corporate interests rather than meeting genuine human needs. This approach results in technology that people must contend with, rather than technology designed to simplify their lives.
Vallor explains that the outcome we receive is something we must then manage, rather than something crafted by us, tailored to our requirements, and intended for our advantage. Additionally it is not the technology we want because instead of customizing technologies to suit human needs, people have to adjust to the demands of technology.
According to the book there are various critical issues in AI development such as systemic biases and privacy threats. Malwina Anna Wójcik notes the exclusion of marginalized groups from the design process, leading to technologies that perpetuate discrimination. Also Matt Malone discusses AI’s impact on privacy, raising concerns about data collection and usage, potentially eroding individual privacy as technology advances.
Consent and knowledge gaps perpetually undermine privacy boundaries. Privacy governs technology’s intrusion into human spheres, yet its definition constantly shifts. Malone concludes that with AI gaining more prominence, privacy will remain crucial in delineating human-technology boundaries.
Experts recommend practical strategies for incorporating a human-centered perspective into AI development including promoting diversity in research, fostering interdisciplinary collaborations, and ensuring transparent data practices.