Bias-reducing legal AI requires ongoing human translators to iterate between models and practice

Evidence

Translators also will play a key role during implementation, studying the workflow to diagnose and troubleshoot problems that arise. Fully integrating transformative legal technology into an office’s processes is a long journey. Machine learning and other techniques that comprise legal technology are more art than science. Successful legal technology requires experimentation, and it is certainly the case that early models will not perform optimally. It also is the case that, once a model is working well, attorneys will struggle to use it properly. In all such instances, translators must highlight what was learned from each prior iteration and present the findings to leadership. — Joseph Avery, Jim Sidanius, Joel Cooper, Gregory Davis, Amanda N., Bias in the Law, 1198

Categorization

• corroborated: 2 independent passages support this claim — Joseph Avery, Jim Sidanius, Joel Cooper, Gregory Davis, Amanda N., Bias in the Law: A Definitive Look at…, 1198