The Emergence of Human-Centered AI

by The Editors

Justin Weisz
Human-Centered AI
3 min readFeb 27, 2023

--

Welcome to our new publication on Human-Centered AI!

Our aim is to publish articles that summarize state-of-the-art research results, case studies, critical reflections, essays, design fictions, and other formats of intellectually stimulating content within the emerging discipline of Human-Centered AI. Please reach out if you’d like to publish with us!

What is Human-Centered AI?

Human-centered AI (HCAI) is an emerging discipline that aims to create AI systems that amplify and augment rather than displace human abilities. The human-centered approach to AI seeks to redress the limitations of traditional approaches to AI that primarily focus on automation and optimizing algorithmic metrics, by preserving human control and by prioritizing well-being, needs and values of individuals, organizations, and the society at large.

The discipline of human-centered AI seeks to ensure that AI systems meet human needs and lead to positive synergy and performance when augmenting human capabilities with AI. Human-centered AI emphasizes the importance of considering the ethical, social, legal, and societal implications of AI systems. Human-centered AI systems:

Research in HCAI intersects with multiple disciplines including human-computer interaction, data visualization, computer-supported-collaborative work, design, psychology, economics, artificial intelligence, and machine learning.

What is this Publication?

This publication serves as an intermediate space for work in human-centered AI. What do we mean by an intermediate space? On one hand, many of us publish our work in academic venues such as CHI, IUI, and NeurIPS. Publications in these spaces typically go through lengthy peer-review cycles, which are important, but may introduce delays in “getting the message out.” On the other hand, many of us also share our work in social media spaces such as Twitter or Mastodon, or in Ben Shneiderman’s newsletters. These spaces work well for short content, but don’t allow for a longer discourse. This publication is intended to serve as a middle ground between these two extremes: a space for publishing thoughtful and detailed articles on hot topics in HCAI, without the lengthy, calendar-bound peer-review process. Here are some examples of what we intend to publish:

  • Deep dives on published or pre-print HCAI papers (excluding papers under peer-review)
  • Essays, critical reflections, or commentary on hot topics in HCAI, or on HCAI as a whole
  • Work that may not be intended for an academic audience or conference publication
  • Recaps of workshops or conferences on HCAI topics (e.g. a summary of HCAI research at NeurIPS, CHI, IUI, etc.)
  • Design fictions depicting possible HCAI futures
  • …and anything else you might think the HCAI community might find intellectually stimulating!

Governance

This publication is governed by a Board of Editors who are responsible for curating, reviewing, and approving all content. The Board of Editors is comprised of human-centered AI researchers with positions spanning academia and industry.

We are the Editors:

What’s Next

There are a few upcoming HCAI workshops to watch out for, and we plan to publish summaries of these events in the coming months.

Learn More & Get Involved

Article updated March 8 to add Upol Ehsan as an editor.

--

--

Justin Weisz
Human-Centered AI

Manager, Senior Research Scientist, & Strategy Lead, Human-Centered AI at IBM Research. My opinions are my own.