Using structured debate between AI systems as a scalable mechanism for humans to judge the quality of AI reasoning.
Organizations
3
Key Papers
2
Grants
3
Total Funding
$85K
First Proposed: 2018 (Irving, Christiano, Amodei)
Cluster: Scalable Oversight
Parent Area: Scalable Oversight
Grants3
| Name | Recipient | Amount | Funder | Date |
|---|---|---|---|---|
| 1-year stipend and compute for conducting a research project focused on AI safety via debate in the context of LLMs. | Paul Bricman | $50K | Long-Term Future Fund (LTFF) | 2022 |
| Developing algorithms, environments and tests for AI safety via debate. | Joe Collman | $25K | Long-Term Future Fund (LTFF) | 2020-07 |
| Upskilling investigation of AI Safety via debate and ML training | Joe Collman | $10K | Long-Term Future Fund (LTFF) | 2019-10 |
Funding by Funder
| Funder | Grants | Total Amount |
|---|---|---|
| Long-Term Future Fund (LTFF) | 3 | $85K |
Tags
debateoversightalignment