Join us on Monday, February 24, 2025, at 12 pm in the Nexus, Hayden Library for a talk led by Zeynep Tufekci, Turkish-American sociologist and Henry G. Bryant Professor of Sociology and Public Affairs at Princeton University.

What if the Real Threat is Artificial Good-Enough Intelligence?

Are we having the wrong nightmares about AI? Many worry that “artificial general intelligence” — AGI, or when machine intelligence matches or exceeds that of humans — poses a severe threat. These newly-superintelligent machines could turn on their creators, we’ve been warned, like Skynet in the apocalyptic movie Terminator. 

But many early predictions — fears and hopes — about how new technologies will change the world turn out to be false or misleading. Meanwhile, significant risks that arise simply from increased speed, expanded scale or lower costs that technology enables are often overlooked. 

A new technology does not have to outperform humans, or even be singularly sensational compared to previous technologies, to fuel extreme turbulence and instability, usually in an unforeseen manner. Cars weren’t so transformative simply because they exceeded the speed benchmark set by horses, nor were they horseless carriages. 

In this talk, Tufekci will examine the disruptive and even potentially catastrophic risks from Artificial Good-Enough Intelligence — AI that can do things not necessarily as well as humans but just good enough to be useful while being faster, cheaper and deployable at scale, and in ways that go beyond current concerns such as employment effects, productivity or bias.