AI adoption requires bilingual role for engineers and users

Evidence

This brings us to the need for translators. Either in-house individuals or collaborating academic researchers must bridge the gap between the data engineers and the employees (often attorneys). These translators must be able to clearly communicate the purpose of the underlying technology. This is distinct from communicating the vision of the project, which falls within leadership’s remit. Translators, by contrast, must be able to explain the intricate details of the technology and why it is producing the output that it is producing. In short, translators must explain the technology to employees, and they must inform data analytics team members of employees’ complaints and suggestions when they arise. — Joseph Avery, Jim Sidanius, Joel Cooper, Gregory Davis, Amanda N., Bias in the Law, 1197

Categorization

Category: Factual Claims

• DEFINE: Defines the ‘translator’ role in AI deployment as a two-way conduit between data engineers and attorney end-users, distinct from leadership’s vision-setting.

Related

• Technical adoption depends on translators bridging engineers and professional users (similarity 0.88)
• Translators are needed to shepherd legal AI adoption through repeated iterations (similarity 0.81)
• Bias-reducing legal AI requires ongoing human translators to iterate between models and practice (similarity 0.80)