Executive summary: People are reluctant to warn about unlikely but potentially catastrophic risks due to fear of reputational damage if the risks do not materialize, but strategies like anonymous warning systems and clarifying risk estimates could help encourage risk warnings.
Key points:
Studies found people are less likely to publicly warn about unlikely risks vs. likely risks, even if the risks have similar expected harm, due to fear of blame if the risks do not occur.
This reluctance to warn was found in the general public, policymakers, and AI researchers. It was even stronger among Chinese participants compared to Americans.
People judge warnings as less justified if the warned risk does not occur (outcome bias), discouraging risk warnings. Even judges and lawyers exhibited this bias.
Anonymous warning systems, routine risk assessment prompts, providing objective evidence, clarifying risk estimates, and overcoming the “first-warner hurdle” are strategies to encourage risk warnings.
Warnings from experts who provide reasons, evidence and intellectual honesty should be seriously considered. Ridiculing sincere warners is reckless and discourages important risk warnings.
This comment was auto-generated by the EA Forum Team. Feel free to point out issues with this summary by replying to the comment, andcontact us if you have feedback.
Executive summary: People are reluctant to warn about unlikely but potentially catastrophic risks due to fear of reputational damage if the risks do not materialize, but strategies like anonymous warning systems and clarifying risk estimates could help encourage risk warnings.
Key points:
Studies found people are less likely to publicly warn about unlikely risks vs. likely risks, even if the risks have similar expected harm, due to fear of blame if the risks do not occur.
This reluctance to warn was found in the general public, policymakers, and AI researchers. It was even stronger among Chinese participants compared to Americans.
People judge warnings as less justified if the warned risk does not occur (outcome bias), discouraging risk warnings. Even judges and lawyers exhibited this bias.
Anonymous warning systems, routine risk assessment prompts, providing objective evidence, clarifying risk estimates, and overcoming the “first-warner hurdle” are strategies to encourage risk warnings.
Warnings from experts who provide reasons, evidence and intellectual honesty should be seriously considered. Ridiculing sincere warners is reckless and discourages important risk warnings.
This comment was auto-generated by the EA Forum Team. Feel free to point out issues with this summary by replying to the comment, and contact us if you have feedback.