Joanna Bryson: How Consciousness Relates to Morality - Why Artificial Consciousness is Unimportant

The Hertie School of Governance

One of the projects of science is to disentangle exactly how many independent factors there are contributing to the workings of the universe we observe. Historically, the state of alertness and capacities of memory we associate with consciousness has been so linked with human moral competence and obligation that in some languages the same words stands for both "conscious" and "conscience." Fortunately, English is not one, so now that we are constructing systems also able to attend to and recall experiences, we can discuss whether this entails anything special with respect to such synthetic systems' moral status. I argue that human morality is rooted in accountability – our capacity to continuously improve our society, and as such depends entirely on our approximately-peer relationship with each other. I discuss how consciousness works in humans and rats, how generative AI like LLMs work and the extent to which such systems can be considered peers and held to account.

 

Guests are welcome!

 

This is the keynote lecture of the Winter School "Ethics of Neuroscience and AI"

organized by John-Dylan Haynes, Marten Kaas, Thomas Schmidt, Dirk Mende, and Lisa Velenosi