Honors News Archive

News in Brief

Are Robots and A.I. an Existential Threat to Humanity? Interdisciplinary Honors students examine the question.

Jul 21, 2016

Will robots take the place of humans as producers of labor and culture?

What will happen to already marginalized workforces around the world as A.I. continues to advance and mechanization reaches all segments of the job market?

Can robots be programmed to empathize and/or be true artists and does it matter if they only appear to love us?

Team Enigma, 2016

These are a few of the discussion topics from Honors 398H, a spring quarter seminar where undergraduates from major departments all across campus convened to consider the course subject: Are Robots and A.I. an Existential Threat to Humanity? 

On the first day of this seminar, UW Honors alumnus and accomplished physicist/educator Rick Freeman ('67) laid out his expectations for the students: "You are not here to reach any particular conclusion, and I am not here to tell you 'the answer' to this question. I would be disappointed, in fact, if we all arrived at the same conclusions. You WILL leave the class with your own unique, informed opinion on this subject and the ability to converse with others whose opinions may differ."

To kick things off, students chose an initial primary position from a six-deep list expressing a range of biases (truncated examples below):

  • The advance of A.I. is highly overrated and will never be an "existential threat"

  • Society will evolve sufficiently fast to accommodate robots in the workplace

  • Regardless of education, we may all be without a self definition of what we do for a living

  • Etc...

A selection of readings and films mixed pop culture speculation with scientific examinations of the subject, weaving lessons from politics, economics, psychology, data analysis, computer science, and the humanities with the storytelling prowess of Hollywood. Viewings of Colossus (released the same year as Dr. Strangelove, reflecting the same societal fears regarding mutually assured nuclear destruction), Her and Ex Machina gave the seminar a cozy feel, students gathered excitedly around a movie screen, munching on pizza. Expert guests included both a visitor from the humanities (Julie Villegas, Honors’ Associate Director and English professor) and a hard sciences rock star (Pedro Domingos, professor of Computer Science & Engineering at UW, author of The Master Algorithm). Professor Domingos said he rarely has a chance to explain machine learning and consider its ramifications with undergraduates whose educational goals and career plans extend beyond the world of computing. Both visits were cited frequently by students in their final course essays.

Punctuating these lessons was a series of small group meetings where students collaborated to refine shared position statements as they proceeded through the course assignments. Team Colossus, Team Human, Team Enigma, and Team Bias presented their findings to the rest of the class and influenced one another as much as the visiting faculty and required readings. Ultimately, each student came to his or her own unique understanding of the question of A.I. and its likely impacts on humanity, distilled into individual essays and final presentations.

Excerpted below from student essays:

I am usually very hesitant to contribute to conversations that disguise a debate, but the other day I participated in a discussion two others were having about A.I. because I felt I had good enough knowledge to contribute. Of course, I have much more to learn and I am no A.I. expert. Knowledge and passion seem to be in a virtuous cycle — each fueling the other. Now that I am passionate about the topic, I will continue to learn more. 

- Savanna Yee

Though human-A.I. collaboration may be feasible in the short term, over time computers will become ever more effective in the work they do, to the point where humans will become distinguished on the job primarily by the errors they make, at which point it will not be economically efficient for them to be employed in such a role. I am convinced that this technological revolution is not like previous ones in human history, in terms of both the sheer pace of change (as deftly compared to the industrialization of the agriculture sector in Humans Need Not Apply), and the type of change (generalizable, self-improving systems that do not create new jobs as quickly as they devour them, as thoroughly evaluated in Rise of the Robots.

Furthermore, I find it more than plausible that the benefits of A.I. will be concentrated largely in the hands of those who previously had the funds to invest in making the most of it, resulting in extreme income inequality, even more so than we see today. The degree to which this is possible, as revealed through research and discussion, also surprised me.

- Kimberly Ruth

Since I believe that the nature of A.I. will be fundamentally different than our own, I do not think humans are at risk from them. That being said, I do fear the concept of an "unintelligent" A.I., as Pedro Domingos warned, given that this type of A.I. might misinterpret what we ask of it or carry out tasks in a harmful manner. But, assuming that we are discussing an A.I. that has some form of logic comparable to that of humans, I believe the only risk we face is in the realm of the job market, where it threatens particularly poorer nations like those in Africa. Other risks are likely present everywhere, but at this point it is impossible to predict what can go wrong. Though I feel much less optimistic about the transition to an automated economy, I still feel that, given the problems we have worked out as a species over the past thousand years, we will be able to reach long-term solutions eventually. 

- Austin Beaulieu

Dr. Freeman closed the seminar praising students for their thoughtful evolution, and inviting them to carry the experience into their priorities for the future. "For those of you in this room — not in the Third World, not in the lower 90% of this country — you won't likely be as impacted as many but you will live to see how it plays out. A.I. will enable you to do much more, but there will be inequality and suffering throughout these transitions. When you see the social unrest, don't just turn your eyes away. See if you can do something about it."