Goals for algorithmic genies


  • Hassan Masum Waterloo Institute for Complexity and Innovation
  • Mark Tovey Western University




Algorithmic genies, artificial intelligence, beneficial AI, goals, ethics, social justice, innovation, foresight, algorithmic boost, raising the floor, Internet.


Algorithmic genies built from growing computational capabilities bring risks like automating well-paying jobs, yet we suggest that if supplied with suitable goals and supporting infrastructure they can help in meeting many human needs. We argue that algorithmic genies can be harnessed to raise the baseline experience of people worldwide (raising the floor), especially if such harnessing is informed by wide consensus and deep evidence. Examples show how algorithmic genies could raise the floor for widely agreed human needs like health, education, and other components of the Social Progress Index. Ensuring that both the least well off and the majority share in the benefits of progress can help to ensure the floor is raised for all (floored progress). Floored progress can apply beyond basic human needs to problems that people across the economic spectrum struggle with (shared floors). We include three tables with illustrative opportunities, and conclude by summarizing the value of raising floors individually and in concert.

Author Biographies

Hassan Masum, Waterloo Institute for Complexity and Innovation

Hassan Masum is a strategist and Affiliate Researcher at the Waterloo Institute for Complexity and Innovation at the University of Waterloo in Canada.

Mark Tovey, Western University

Mark Tovey is a Postdoctoral Fellow in the Department of History and in the Centre for Planetary Science and Exploration at Western University in Canada. He is the editor of Collective Intelligence: Creating a Prosperous World at Peace, and co-editor (with Hassan Masum) of The Reputation Society: How Online Opinions Are Reshaping the Offline World.




How to Cite

Masum, H., & Tovey, M. (2018). Goals for algorithmic genies. First Monday, 23(2). https://doi.org/10.5210/fm.v23i2.8073