87 - AI and the Value Alignment Problem

12/23/20 • -1 min

How do we make sure that an AI does the right thing? How could we do this when we ourselves don't even agree on what the right thing might be? In this episode, I talk to Iason Gabriel about these questions. Iason is a political theorist and ethicist currently working as a Research Scientist at DeepMind. His research focuses on the moral questions raised by artificial intelligence. His recent work addresses the challenge of value alignment, responsible innovation, and human rights. He has also been a prominent contributor to the debate about the ethics of effective altruism.

You can download the episode here or listen below. You can also subscribe on Apple Podcasts, Stitcher, Spotify and other podcasting services (the RSS feed is here).

Show Notes:

Topics discussed include:

What is the value alignment problem?
Why is it so important that we get value alignment right?
Different ways of conceiving the problem
How different AI architectures affect the problem
Why there can be no purely technical solution to the value alignment problem
Six potential solutions to the value alignment problem
Why we need to deal with value pluralism and uncertainty
How political theory can help to resolve the problem

Relevant Links

Iason on Twitter
"Artificial Intelligence, Values and Alignment" by Iason
"Effective Altruism and its Critics" by Iason
My blog series on the above article
"Social Choice Ethics in Artificial Intelligence" by Seth Baum

Subscribe to the newsletter

You can download the episode here or listen below. You can also subscribe on Apple Podcasts, Stitcher, Spotify and other podcasting services (the RSS feed is here).

Show Notes:

Topics discussed include:

What is the value alignment problem?
Why is it so important that we get value alignment right?
Different ways of conceiving the problem
How different AI architectures affect the problem
Why there can be no purely technical solution to the value alignment problem
Six potential solutions to the value alignment problem
Why we need to deal with value pluralism and uncertainty
How political theory can help to resolve the problem

Relevant Links

Iason on Twitter
"Artificial Intelligence, Values and Alignment" by Iason
"Effective Altruism and its Critics" by Iason
My blog series on the above article
"Social Choice Ethics in Artificial Intelligence" by Seth Baum

Subscribe to the newsletter

Previous Episode

86 - Are Video Games Immoral?

Have you ever played Hitman? Grand Theft Auto? Call of Duty? Did you ever question the moral propriety of what you did in those games? In this episode I talk to Sebastian Ostritsch about the ethics of video games. Sebastian is an Assistant Prof. (well, technically, he is a Wissenschaftlicher mitarbeiter but it's like an Assistant Prof) of Philosophy based at Stuttgart University in Germany. He has the rare distinction of being both an expert in Hegel and the ethics of computer games. He is the author of Hegel: Der Welt-Philosoph (published this year in German) and is currently running a project, funded by the German research body DFG, on the ethics of computer games.

You can download the episode here or listen below. You can also subscribe on Apple Podcasts, Stitcher, Spotify and other podcasting services (the RSS feed is here).

Show Notes

Topics discussed include:

The nature of video games
The problem of seemingly immoral video game content
The amorality thesis: the view that playing video games is morally neutral
Defences of the amorality thesis: it's not real and it's just a game.
Problems with the 'it's not real' and 'it's just a game' arguments.
The Gamer's Dilemma: Why do people seem to accept virtual murder but not, say, virtual paedophilia?
Resolving the gamer's dilemma
The endorsement view of video game morality: some video games might be immoral if they endorse an immoral worldview
How these ideas apply to other forms of fictional media, e.g. books and movies.

Relevant Links

Sebastian's homepage (in German)
Sebastian's book Hegel: Der Weltphilosoph
'The amoralist's challenge to gaming and the gamer's moral obligation' by Sebastian
'The immorality of computer games: Defending the endorsement view against Young’s objections' by Sebastian and Samuel Ulbricht
The Gamer's Dilemma by Morgan Luck
Homo Ludens by Johan Huizinga

Subscribe to the newsletter

Next Episode

88 - The Ethics of Social Credit Systems

Should we use technology to surveil, rate and punish/reward all citizens in a state? Do we do it anyway? In this episode I discuss these questions with Wessel Reijers, focusing in particular on the lessons we can learn from the Chinese Social Credit System. Wessel is a postdoctoral Research Associate at the European University Institute, working in the ERC project “BlockchainGov”, which looks into the legal and ethical impacts of distributed governance. His research focuses on the philosophy and ethics of technology, notably on the development of a critical hermeneutical approach to technology and the investigation of the role of emerging technologies in the shaping of citizenship in the 21st century. He completed his PhD at the Dublin City University with a Dissertation entitled “Practising Narrative Virtue Ethics of Technology in Research and Innovation”. In addition to a range of peer-reviewed articles, he recently published the book Narrative and Technology Ethics with Palgrave, which he co-authored with Mark Coeckbelbergh.

You can download the episode here or listen below.You can also subscribe on Apple Podcasts, Stitcher, Spotify and other podcasting services (the RSS feed is here).

Show Notes

Topics discussed in this episode include

The Origins of the Chinese Social Credit System
Historical Parallels to the System
Social Credit Systems in Western Cultures
Is China exceptional when it comes to the use of these systems?
The impact of social credit systems on human values such as freedom and authenticity
How the social credit system is reshaping citizenship
The possible futures of social credit systems