Course page Home page Home pageNew course

AI safety | Zoonk

71. AI safety

Identify failure modes such as reward hacking, deceptive outputs, uncontrolled actions, and loss of human control, then apply capability evaluations and layered safeguards.

Artificial Intelligence

Create this chapter

This chapter is part of the course, but it hasn't been created yet. Create it to see all of its lessons.

Create chapterNBack to courseEsc