Podcast Club ONLINE: AI-Safety Expert's Views and Predictions on Our Future

This week's podcast is DOAC (The Diary of a CEO) with host Steven Bartlett and guest Roman Yampolskiy, a computer scientist at the University of Louisville, who is mostly known for his work on AI safety and cybersecurity. Yampolskiy is considered a leading voice in AI safety. He is the author of five books on AI and is credited with over 100 publications, primarily on AI safety, cybersecurity, and digital forensics. Yampolskyi is believed to have coined the term "AI safety" in a 2011 publication and is an early researcher in the field. He is the founder and former director of the Cyber Security Lab in the Department of Computer Engineering and Computer Science at the University of Louisville's Speed School of Engineering.

PLEASE NOTE: We require that attendees listen to the podcast below before attending the meeting to optimize discussion.

https://podcasts.apple.com/us/podcast/roman-yampolskiy-these-are-the-only-5-jobs-that-will/id1291423644?i=1000724933846

or

https://www.youtube.com/watch?v=UclrVWafRAI

(1 hr, 29 min.)

-------------

What's this podcast about?
Frankly, this week's podcast is very grim, unsettling, and disturbing. We all live in a world where AI is becoming ubiquitous. Whether we desire it or not, unchecked AI "improvements" are being implemented constantly, and we are all hurling toward an unknown future that has the potential of changing humanity at a pace that far exceeds anything ever experienced before. What risks might we be confronting? Are all the AI doomsday predictions just absurdly alarmist and attention-grabbing? Or, given that many serious and supposedly-knowledgeable AI experts are increasingly sounding the alarm about AI safety and that all this has existential consequences, should we in fact be truly alarmed?
This podcast presents the doomsday (though potentially accurate) view. Again, it is disturbing and thus, may not be for you. In this episode, Yampolskiy gives his predictions for the future using specific years by which he expects certain outcomes (assuming no reining in of AI development) and explains the basis for his beliefs. He also discusses what he expects will happen to the job market as AI becomes able to do things better, faster, and cheaper than humans. Yampolskiy addresses the "Can't we just unplug it?" question and ultimately gives his solution on how to use AI (which he favors) in a way that won't jeopardize humanity. He also talks about whether we live in a simulation, and for some (including me), this may be seem a bit too far out there. Nevertheless, Yampolskiy's voice is one of many that are truly sounding the alarm on unchecked AI development, so there's much food for thought here.

Zoom link: https://us02web.zoom.us/j/3682869318?pwd=SC80SzI1S3JkeW1pbGhzcTNVQWVmdz09. Password for event might be W5hDy7

What's a podcast club?
It's like a book club for podcasts. Each week we discuss an interesting and currently-relevant episode from podcasters like Ezra Klein, Michael Shermer, or Russ Roberts. We start off as a large group and spend most of the time talking in small break-out groups, which we remix 3 times during the course of the evening. Our conversations are casual and open. We ask that everyone speak respectfully at all times, and we encourage free and concise discussion relevant to the week's podcast topic. To do this, we try to adhere to Grice's Maxims: https://bit.ly/2p4uSQm

Ultimately, the idea is to help each other think a little differently, learn a thing or two and have some fun.