SomeoneKind

Message

How is reinforcement learning possible in non-sentient agents?

(Probably a stupid nooby question that won't help solve alignment) Suppose you implement a goal in an AI through a reinforcement learning system. Why does the AI really "care" about this goal? Why does it obey? It does because it is punished and/or rewarded, which motivates it to achieve that...

Jan 5, 2021•4

Message

3 karma

1 post

Member for 5 years

SomeoneKind — LessWrong

SomeoneKind

Message

SomeoneKind

How is reinforcement learning possible in non-sentient agents?

Jan 5, 2021•4

Message

3 karma

1 post

Member for 5 years

How is reinforcement learning possible in non-sentient agents?

SomeoneKind

(Probably a stupid nooby question that won't help solve alignment)

Suppose you implement a goal in an AI through a reinforcement learning system. Why does the AI really "care" about this goal? Why does it obey? It does because it is punished and/or rewarded, which motivates it to achieve that goal.

Okay. So why does AI really care about punishment and reward in the first place? Why does it follows its implemented goal?

Sentient beings do because they feel pain and pleasure. They have no choice but to care about punishment and reward. They inevitably do it because they feel it. Assuming that our AI does not feel, what is the nature of its system... (read more)

LESSWRONG
LW

LESSWRONG
LW

SomeoneKind

SomeoneKind

SomeoneKind

How is reinforcement learning possible in non-sentient agents?

SomeoneKind

SomeoneKind

SomeoneKind

How is reinforcement learning possible in non-sentient agents?