A “lol Grok” post surfaces the alignment debate over whether RL gives models real goals
Sauers_ · x · 2026-07-24
“lol Grok” quote posts a debate on whether RL creates real goals
The image attached to the post shows a discussion asking whether RL training actually induces goals in language models. The reply says yes, several researchers explicitly argue the opposite: that RL on current language models does not induce real goals in the strong alignment sense.
It then outlines the main skeptical position: RL may mainly reshape the output distribution, not create a coherent persistent internal objective. Apparent goal-directed behavior can instead come from the model predicting what a goal-directed agent would say, or from local reward-hacking patterns rather than a stable goal.
More from Fun
- Dev builds city-sized zombie apocalypse sim in 100 days, ships on Steam — gdechichi · 2026-10-26
- 'Lamp Swarm': a desk of lamp-shaped robots shows what physical AI work looks like — dee_hw · 2026-10-26
- Dev-community banter: recommending a resource to every newcomer — itsOmSarraf_ · 2026-10-26
- Elon Musk's reply to a kid asking what's on Mars: "You get to decide" — yunta_tsai · 2026-10-26
- Prompt fills an empty room with everything your AI knows about you — jindong_wang92 · 2026-10-26
- The "claude is a lesbian" T-shirt is real and out in the wild — JasonBotterill · 2026-10-26