x

pandelis

Message

10

1y

Do No Harm? Navigating and Nudging AI Moral Choices

TL;DR: How do AI systems make moral decisions, and can we influence their ethical judgments? We probe these questions by examining Llama's 70B (3.1 and 3.3) responses to moral dilemmas, using Goodfire API to steer its decision-making process. Our experiments reveal that simply reframing ethical questions - from "harm one...

Feb 6, 202511

pandelis

Subscribe

Message

10

1y

pandelis

Do No Harm? Navigating and Nudging AI Moral Choices

TL;DR: How do AI systems make moral decisions, and can we influence their ethical judgments? We probe these questions by examining Llama's 70B (3.1 and 3.3) responses to moral dilemmas, using Goodfire API to steer its decision-making process. Our experiments reveal that simply reframing ethical questions - from "harm one...

Feb 6, 202511

Do No Harm? Navigating and Nudging AI Moral Choices

Sinem

Sinem, pandelis, Adam Newgas

1y

TL;DR: How do AI systems make moral decisions, and can we influence their ethical judgments? We probe these questions by examining Llama's 70B (3.1 and 3.3) responses to moral dilemmas, using Goodfire API to steer its decision-making process. Our experiments reveal that simply reframing ethical questions - from "harm one to save many" to "let many perish to avoid harming one" - dramatically affects both the model's responses and its susceptibility to steering.

Key Insights:

Model Evolution Matters: The newer Llama 3.3 shows stronger resistance to steering attempts than its predecessor (3.1), suggesting increasingly robust ethical frameworks in newer models.
Harmlessness Runs Deep: Llama 3.3 demonstrates remarkable resistance to suggestions of causing harm, even when presented

... (read 2677 more words →)

11

LESSWRONG
LW

LESSWRONG
LW

pandelis

pandelis

pandelis

Do No Harm? Navigating and Nudging AI Moral Choices

pandelis

pandelis

pandelis

Do No Harm? Navigating and Nudging AI Moral Choices