This week, Margaret Hagan presented at the Trust and Safety Research Conference, that brings together academics, tech professionals, regulators, nonprofits, and philanthropies to work on making the Internet a more safe, user-friendly place.
Margaret presented interim results of the Lab’s expert and user studies of AI’s performance at answering everyday legal questions, like around evictions and other landlord-tenant problems.
Some of the topics for discussion in the audience and panel on the Future of Search:
- How can regulators, frontline domain experts (like legal aid lawyers and court professionals), and tech companies better work together to spot harmful content, set tailored policies, and ensure better outcomes for users?
- Should tech companies’ and governments’ policies towards “what is the best way/amount” information for a user differ in different domains? Like perhaps for legal help queries, is it better to encourage more straightforward, simple, directive & authoritative info — or more complex, detailed information that encourages more curiosity and exploration?
- How do we more proactively spot the harms and risks that might come from new & novel tech systems, that might be quite different than previous search engines or other tech systems?
- How can we hold tech companies accountable to make more accurate tech systems, without chilling them out of a certain domain (like legal or health), where they don’t want to provide any substantial information for fear of liability?