Executive summary: Discovering ways for companies to align AI systems with social good that also improve profitability reduces AI risk by shaping companies’ incentives.
Key points:
Alignment taxes like robustness tradeoffs incur costs for safety, while alignment windfalls like human feedback also create value.
Startups optimize greedily within the landscape of taxes and windfalls they know of to maximize returns.
We can shape this landscape via regulation, public awareness, recruiting appeals.
But companies’ knowledge of the full landscape is limited, so promoting windfalls specifically also guides development.
“Factored cognition” improves transparency and trust while solving problems, an example of a windfall.
Discovering and advocating for more windfalls makes AI safety intrinsically economically rational.
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: Discovering ways for companies to align AI systems with social good that also improve profitability reduces AI risk by shaping companies’ incentives.
Key points:
Alignment taxes like robustness tradeoffs incur costs for safety, while alignment windfalls like human feedback also create value.
Startups optimize greedily within the landscape of taxes and windfalls they know of to maximize returns.
We can shape this landscape via regulation, public awareness, recruiting appeals.
But companies’ knowledge of the full landscape is limited, so promoting windfalls specifically also guides development.
“Factored cognition” improves transparency and trust while solving problems, an example of a windfall.
Discovering and advocating for more windfalls makes AI safety intrinsically economically rational.
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.