Research on building AI systems that allow themselves to be corrected, modified, or shut down by human operators.
Organizations
3
Key Papers
2
Grants
2
Total Funding
$105K
First Proposed: 2015 (Soares et al., MIRI)
Cluster: Scalable Oversight
Parent Area: Scalable Oversight
Grants2
| Name | Recipient | Amount | Funder | Date |
|---|---|---|---|---|
| AI Alignment Awards — Shutdown Problem Contest | AI Alignment Awards | $75K | Coefficient Giving | 2022-09 |
| Building towards a "Limited Agent Foundations" thesis on mild optimization and corrigibility | Alex Turner | $30K | Long-Term Future Fund (LTFF) | 2019-04 |
Funding by Funder
| Funder | Grants | Total Amount |
|---|---|---|
| Coefficient Giving | 1 | $75K |
| Long-Term Future Fund (LTFF) | 1 | $30K |
Tags
corrigibilityshutdownsafety-research