Instead of my current academic work on AI ethics, I'll present on a recent satirical novel about science I wrote, The Strength of the Illusion. I'll do a short reading from the book on a section which expounds on values in AI. I'll use this as a jumping off point for us to have an informal broader conversation about the capacities of current language models with connections to issues such as on consciousness, morality, and the money behind the science.
Here is a short description of the book:
A timely meditation on what happens when artificial intelligence clashes with human stupidity.
An AI researcher, Ty, has discovered how to teach a machine to write. He joins a start-up, Opel, eager to bring on-demand literature to millions. As Opel makes overbold claims about how its writing machine will automate human connection, Ty is increasingly drawn to the fiery relationship with his activist partner, Zora. As each of them flees from their own past, Ty and Zora while away hours in their passionate debates about how to create a future for themselves and for the world. When Zora urges Ty to join her protest against big tech, Ty is forced to decide what he really values. Caught between worlds, Ty loses himself in the advice of his writing machine. How will Ty come to understand the difference between right and wrong, real and illusory?
You can read more about the book here:
Jared Moore is a writer and an AI researcher thinking about how people and computers work and don’t work. Currently, his work focuses on social reasoning. He is a PhD student in computer science at Stanford University. Most recently, he was a lecturer at the University of Washington where he taught classes on the math, philosophy, and ethics of AI. He has worked at AI2, Xnor.ai, and WadhwaniAI.