Our goal is to be reliable on any text that seems like kind of normal human text. That's a really neat problem. Yeah in a sense you're trying to find at least one case where adversarial examples don't work. It seems like you can kind of construct an adversarial example from the bottom up. You keep mutating it a little more so the classifier gets a little more wrong and a little morewrong. And then see which way pushes it towards the wrong label the most.
Read the full transcript here.
How hard is it to arrive at true beliefs about the world? How can you find enjoyment in being wrong? When presenting claims that will be scrutinized by others, is it better to hedge and pad the claims in lots of caveats and uncertainty, or to strive for a tone that matches (or perhaps even exaggerates) the intensity with which you hold your beliefs? Why should you maybe focus on drilling small skills when learning a new skill set? What counts as a "simple" question? How can you tell when you actually understand something and when you don't? What is "cargo culting"? Which features of AI are likely in the future to become existential threats? What are the hardest parts of AI research? What skills will we probably really wish we had on the eve of deploying superintelligent AIs?
Buck Shlegeris is the CTO of Redwood Research, an independent AI alignment research organization. He currently leads their interpretability research. He previously worked on research and outreach at the Machine Intelligence Research Institute. His website is shlegeris.com.
Staff
Music
Affiliates