| |
|
| |
What does it mean for a machine to be intelligent, & can we define it mathematically?
Join 90/30 for a reading group on Shane Legg's Machine Super Intelligence. This 2008 PhD thesis explores universal artificial intelligence, the theoretical agent AIXI, & the limits of building machines that learn & act across unfamiliar environments.
We'll discuss:
How can we measure intelligence beyond performance on a single task?
What does AIXI tell us about an ideal intelligent agent?
What changes when we move from theoretical intelligence to systems with limited compute?
How useful are these ideas for understanding today's AI?
This is a long read, so come with a section, question, or argument you want to unpack. You don't need to understand every proof to join the conversation. We are only mandating chapter 1 & 4 while 2 is suggested.
|
|
|
|
|
|
|
|
