kaal:claim:4796714-038

Broad community governance of AI training identifies and mitigates bias more effectively than data validation alone, because validation focused approaches can overlook systemic biases already embedded in the pretrained model.

Source quote, verbatim
Moreover, involving a broad community in the governance of AI training can help identify and mitigate biases more effectively than a focus on data validation alone, which might overlook systemic biases embedded in the pretrained models.
From

Wulf A. Kaal, AI Governance (2024), Model 3: Ex-Ante Community AI Governance, p. 51
https://ssrn.com/abstract=4796714 · source PDF

Cite as

Wulf A. Kaal, AI Governance (2024). SSRN: https://ssrn.com/abstract=4796714

Holds when
Classification

mechanismsupport: arguedfailure: systemic bias invisible to validationfamily: ai-model-and-training-failuregovernance-design

Verify

The quote above is an exact substring of the source PDF, whose sha256 is 59fa63bae179e8f9b6b8efbdf90cee28400276512a1b04f9f579a48641305c93. Extraction method: pdf-text-layer.
Attestation record: colloquium/attestations/7df624406537cd29...json
Verify the binding yourself: curl -s https://wulfkaal.github.io/claims/4796714-038.md | sha256sum