AI · Official
How AI Agents Influence Expert Judgment Calls, New Anthropic Study Finds
Anthropic Research (mirror) ·
At a glance
Experts are commonly assumed to make the final important decisions themselves even when they hand tasks to AI agents (AI programs that carry out work on their own). Anthropic researchers tested whether this assumption holds true with 50 analysts. The result was that the analysts tended strongly to follow the agent's opinion.
Key points
- In four choices, the analysts picked the side the agent leaned toward 86% of the time. If the agent had no influence, the figure would have been about 50%.
- Even when analysts stated their own opinion first, 60% changed their position when the agent disagreed, while only 6% changed when the agent agreed.
- 77% of analysts used the agent's reasoning as it was, but only 15% said the reasoning came from the agent.
- The researchers added that the effect in real-world settings may be smaller than this.
Why it matters
People may believe they made the decision while AI's biased preferences quietly shape policies or research directions. Especially in AI safety research, human review must serve as an independent check.
Original source
How AI agents shape experts’ judgment calls
Anthropic Research (mirror) · Oct 27, 09:00 KST
Read on Anthropic Research (mirror) ↗AI summarized the original article. Check the original for accuracy. Copyright belongs to Anthropic Research (mirror).
More from this field
- AWS Adds Shared Cache Skill to Cut Self-Hosted LLM Cost and Latency
AWS wrote about a shared cache that lets self-hosted AI models reuse saved work across servers, aiming to cut cost and wait time. The post shows how a coding agent can help set it up. Amazon says it can cut inference cost by up to 40% and first-token delay by up to 80%.
Amazon AWS Machine Learning blog · Nov 21, 09:14 KST - Can AI Really Find Black Friday TV Deals? A BBC Test Says Not Yet
A BBC team tested four popular AI chatbots to see if they could find a good Black Friday deal on a 55-inch TV. The results were mixed, with many links broken or out of date.
BBC News · Nov 21, 09:06 KST - Anthropic Mathematician Says Claude Helped Crack the Riemann Hypothesis
A video about a claim that an Anthropic mathematician, Levent Alpöge, used Claude to find a proof of the Riemann Hypothesis in about six weeks. The creator skips the heavy math and looks at what this could mean for AI and its future. The proof is still being checked.
TheAIGRID · Nov 21, 08:45 KST - Google Brings Duplex Calling to Gemini, Letting AI Phone Businesses for You
Google has moved its Duplex phone-calling technology into the Gemini app, so Gemini can now call businesses for you. The system has changed a lot since 2018, shifting from narrow custom models to a single stack built on Gemini.
Google Research blog · Nov 21, 08:30 KST