Person
Jacob Steinhardt
cited by eight posts · in two sources · first cited September 2026
In their words
“Steinhardt (2023) outlines a number of reasons to expect LLMs to use these skills to optimize for achieving specific outcomes, and surveys cases in which existing LLMs adopt goal-directed “personas”.”
· · machine-resolved
“McCandlish, Luke Muehlhauser, Richard Ngo, David Roodman, Rohin Shah, Carl Shulman, Nate Soares, Jacob Steinhardt, and Eliezer Yudkowsky for input on the longer report on which this essay is based; thanks to Leopold”
· Existential Risk from Power-Seeking AI · machine-resolved