Ethical Machines

Innovation is great…but hype is bad. Not only has all this talk of innovation not increased innovation, but it also creates a bad environment in which leaders can make reasoned judgments about where to devote resources. So says Lee Vinsel in my latest podcast episode.

ALSO: We want proactive regulations before the sh!t hits the fan, right? Not so fast, says Lee. Proactive regulations presuppose we’re good at predicting how technologies will be applied, and we have a terrible track record on that front. Perhaps reactive regs are appropriate (and we need to focus on making a more agile government).

Super interesting conversation that will push you to think differently about innovation and what appropriate regulation looks like.

Lee Vinsel is an Associate Professor of Science, Technology, and Society at Virginia Tech and host of Peoples & Things, a podcast about human life with technology. His work examines the social dimensions of technology with particular focus on the relationship between government and technological change. He is the author of Moving Violations: Automobiles, Experts, and Regulations in the United States and, with Andrew L. Russell, The Innovation Delusion: How Our Obsession with the New Has Disrupted the Work That Matters Most.

Digital twins: they're not just a sci-fi doppelganger—they're a game-changing technology that can simulate real-world scenarios in real-time. My latest chat with Ingrid Vasiliu-Feltes opened my eyes to the Pandora's Box of ethics we're cracking open. 

It's a moral labyrinth. I went from "Why should I care?" to "Oh, I really SHOULD care," and trust me, you will too.

You might think it's outrageous that companies collect data about you and use it in various ways to drive profits. The business model of the "attention" economy is often objected to on just these grounds.

On the other hand, does it really matter if data about you is collected and no person ever looks at that data? Is that really an invasion of your privacy?

Carissa and I discuss all this and more. I push the skeptical line, trying on the position that it doesn't really matter all that much. Carissa has powerful arguments against me.

This conversation goes way deeper than 'privacy good/data collection bad' statements we see all the time. I hope you enjoy!

Carissa Véliz is an Associate Professor in Philosophy at the Institute for Ethics in AI, and a Fellow at Hertford College at the University of Oxford. She is the recipient of the 2021 Herbert A. Simon Award for Outstanding Research in Computing and Philosophy. She is the author of the highly-acclaimed Privacy Is Power (an Economist book of the year, 2020) and the editor of the Oxford Handbook of Digital Ethics. She advises private and public organisations around the world on privacy and the ethics of AI.

Job automation, human creativity, and generative AI in higher education, all wrapped into one. Questions include:

Will there be fewer jobs for designers because gen AI will create marketing materials, websites, etc.?

Will cameras go the way of the dark room?

What’s the role of Gen AI in fine art?

What do art teachers in higher Ed do about the new tool?

As an artist and faculty at the School of Visual Arts, Eric is in a rare position to have insight into all of this. And he’s been my closest friend for the last 25 years :)

Eric Corriel is a multidisciplinary artist living in New York City. After graduating from Cornell University with a Bachelor of Arts in Philosophy, he went on to get a Diplôme National d’Arts Plastiques from the École Régionale Supérieure d’Expression Plastique in Tourcoing, France. Currently living in New York City, Eric takes the urban landscape as a medium in which to create site-specific installations. He also teaches Artist as Activist at School of Visual Arts in New York City, where he is also Digital Strategy Director.

Eric is a two-time New York State Council on the Arts grant recipient, two-time Webby Award winner, and New York Foundation of the Arts Fellow

Humans are bad at making predictions, especially in a criminal justice setting. And it looks like AI can do better both from an accuracy and bias standpoint. So let’s replace human judges with AI. So argues professor of law Peter Salib in our fascinating discussion.

Peter Salib is an Assistant Professor of Law at the University of Houston Law Center and Associated Faculty in the Hobby School of Public Affairs. He writes and teaches about law and artificial intelligence. His scholarly work has been published in, among others, The University of Chicago Law Review, Northwestern University Law Review, Texas Law Review, and the Duke Law Journal Online. Before joining the University of Houston Law Center, Peter was a Climenko Fellow at Harvard Law School and a judicial clerk for the Honorable Frank H. Easterbrook. Before that, he practiced law at Sidley Austin, LLP, specializing in appellate litigation.

I talk a lot about bias, black boxes, and privacy, but perhaps my focus is too narrow. In this conversation, Aimee and I discuss what she calls “sustainable AI.” We focus on the environmental impacts of AI, the ethical impacts of those environmental impacts, and who is paying the social cost of those who benefit from AI.

Aimee van Wynsberghe is the Alexander von Humboldt Professor for Applied Ethics of Artificial Intelligence at the University of Bonn in Germany. Aimee is director of the Institute for Science and Ethics and the Bonn Sustainable AI lab. She is co-director of the Foundation for Responsible Robotics and a member of the European Commission's High-Level Expert Group on AI. She is a founding editor for the international peer-reviewed journal AI & Ethics and member of the World Economic Forum's Global Futures Council on Artificial Intelligence and Humanity. She is author of the book Healthcare Robots: Ethics, Design, and Implementation and is regularly interviewed by media outlets. In each of her roles, Aimee works to uncover the ethical risks associated with emerging robotics and AI. Aimee’s current research, funded by the Alexander von Humboldt Foundation, brings attention to the sustainability of AI by studying the hidden environmental costs of developing and using AI. 

Is it better to have a high performing black box AI or a lower performing explainable AI?

Are the explanations for how AI works actually true or really a distortion of what's going on inside the model?

How should we think about and operationalize the tradeoffs between running explainability algos and the high cost and carbon footprint of running them?

These questions and more with Kristof. A really fascinating discussion that reveals the complexity behind simplistic calls for explainable AI.

I doubt there’s a large corporation out there that hasn’t been pitched at least a dozen or so AI tools for HR. From vetting resumes to hiring to promoting to firing to predicting the likelihood someone will quit, there’s an AI tool for that.

But HR usually doesn’t know how to vet these systems. Nor does the standard procurement process. And businesses almost never have a process by which HR or procurement can hand these things over to internal AI ethical risk experts.

What’s more, the idea that we can have independent parties “audit” the algorithms for bias is a gross oversimplification of what needs to happen.

I talk about all this and more with Hilke Schellmann and Mona Sloane, Ph.D., both of whom know way more than I do about the ways AI stands between people and the jobs they need.

Hilke Schellmann is an Emmy-award-winning journalism professor at New York University and a freelance reporter holding artificial intelligence accountable. Her work has been published in The Wall Street Journal, The Guardian, The New York Times, and MIT Technology Review, among others. She is currently writing a book on artificial intelligence and the future of work for Hachette.

Mona Sloane, Ph.D. is a sociologist working on design and inequality, specifically in the context of AI design and policy. She is a Research Assistant Professor at NYU’s Tandon School of Engineering, Senior Research Scientist at the NYU Center for Responsible AI, a Fellow with NYU’s Institute for Public Knowledge (IPK) and The GovLab, and the Director of the *This Is Not A Drill* program on technology, inequality and the climate emergency at NYU’s Tisch School of the Arts. She is the principal investigator on multiple research projects on AI and society, and holds an affiliation as postdoctoral scholar with the Tübingen AI Center at the University of Tübingen in Germany where she leads a 3-year federally funded research project on the operationalization of ethics in German AI startups. Mona founded and runs the IPK Co-Opting AI series at NYU and currently serves as editor of the technology section at Public Books. She holds a Ph.D. in Sociology from the London School of Economics and Political Science. Follow her on Twitter @mona_sloane.

You think targeted marketing can manipulate users and populations? Just wait.

Imagine chatbots powered by LLMs at scale. We'll see chatbots trained to be the best salesperson, negotiator, and manipulator you've ever encountered.

All this and more in my conversation with Louis Rosenberg.

Dr. Louis Rosenberg is a longtime technologist in the fields of augmented reality, virtual reality and artificial intelligence. His work began over thirty years ago in labs at Stanford and NASA. In 1992 he developed the first mixed reality system at Air Force Research Laboratory. In 1993 he founded the early VR company Immersion Corporation which he brought public on NASDAQ. In 2004 he founded Outland Research to develop AR technology that was acquired by Google in 2011. And in 2014 he founded Unanimous AI to amplify the intelligence of human groups using the biological principle of Swarm Intelligence. Rosenberg received his PhD from Stanford University, was a tenured professor at California State University, and has been awarded over 300 patents for VR, AR, and AI technologies. He's currently CEO of Unanimous AI, the Chief Scientist of the Responsible Metaverse Alliance, and the Global Technology Advisor to the XR Safety Initiative.