Event

Philosophy Talk: Rethinking Artificial General Intelligence: Beyond Anthropomorphism

Date, time, location

Date

Friday, April 11, 2025

Time

4:30 pm to 5:30 pm EDT

Location

King Building, 343

10 N. Professor St.
Oberlin, OH 44074

Get Directions

Additional details

Cost

Free

Join the Oberlin College Philosophy Department guest speaker Ben Baker, JD PhD, Assistant Professor, Colby College Department of Philosophy, for the talk Rethinking Artificial General Intelligence: Beyond Anthropomorphism.

Abstract: 

Advances in AI technology have led many to predict the arrival of artificial general intelligence (AGI), but how should AGI be defined? Often AGI is conceptualized in relation to human intelligence, and in this talk I will challenge this anthropomorphic approach and propose an alternative view. I'll begin by clarifying the problem, noting that intelligence must be understood in relation to a particular space of problems. I'll then critically assess the two main approaches to defining AGI anthropomorphically: one based on comparing biological forms of intelligence, and one focused on tasks we have designed machines for. The biological approach confronts a "problem of unnatural selection" arising from essential differences in evolutionary selection processes and the analogous selection processes in AI development. The task-centered approach fails to consider three key domains of human intelligence: self-modification, emotion, and social interaction. Thus, while AI is becoming more capable of solving a wider range of problems that matter to us, it is not converging toward a human-like form of intelligence. In place of an anthropomorphic view, I propose a value-oriented framework that defines AGI through four interconnected spheres of human concern, which I call the "Four Ps": Preservation (ecological and economic sustainability), Progress (knowledge and technological innovation), Person (individual well-being), and Politics (social justice and equitable power relations). This anthropocentric rather than anthropomorphic approach provides a clearer vision for how increasingly general AI can address our most pressing challenges, while acknowledging that AGI will likely possess fundamentally different characteristics than human intelligence.


Open to all members of the Oberlin campus community

Event Contact

Martin Thomson-Jones

440-775-8390

mjones@oberlin.edu

You may also like…

Trauma AWARE (101) Workshop

Details: Date, Time, and Location

  • Time

    12:00 pm to 1:30 pm EDT
  • Location

    RSVP for in-person location