This website is developed by Ben Shneiderman (ben@cs.umd.edu) and Mengnan Du (mengnan.du@njit.edu) to support the growing community of those who believe in Human-Centered AI. Further information at https://hcil.umd.edu/human-centered-ai/ and at the Google Group on Human-Centered AI https://groups.google.com/g/human-centered-ai/about or follow us on Twitter @HumanCenteredAI
The rapidly emerging interest in Human-Centered AI is based on at least these design goals: auditable, comprehensible, dependable, ethical, explainable, fair, humane, interpretable, predictable, reliable, resilient, responsible, robust, safe, testable, transparent, trustworthy, unbiased, usable, and verifiable.
The advocates of Human-Centered AI promote human values such as rights, justice, and dignity. They seek to design, implement, and disseminate supertools that support human self-efficacy, creativity, responsibility, and social connections. These supertools will be reliable, safe, and trustworthy systems even in the face of threats from malicious actors, biased data, and flawed software. Thoughtful design strategies can deliver high levels of human control and high levels of automation, as they do already in digital cameras, navigation tools, and much more. The future will be shaped by those who support human autonomy, well-being, and control over emerging technologies. The future is human-centered.