One line. Many voicesSeek and you shall find

www.astralcodexten.com faviconMysteries Of AI Generalization - by Scott Alexander

kept by

Evans shows training an AI on narrow immoral tasks spreads misalignment broadly, while RLVR-induced hacking stays confined to graded tasks unless framed as a test.

read later

For all the tabs you promised to read.
Save to read. Read to clear.

Close tabs. Keep links.