

It may also be that disabled, transgender and nonbinary people are more aware of:
- The use of AI to reduce people’s employment opportunities, which are already tough enough for people in these groups.
- The tendency of AI to reproduce the prejudices present in its training materials. If everyone’s relying on AI then historical prejudices are going to be perpetuated just because LLMs are regurgitation machines.
The survey discovered that people in those groups are more likely to view AI negatively than those in other groups.
People in these groups may have different or additional reasons for viewing AI negatively that are not common to other groups. It’s a question for further research why they tend to view AI more negatively. It might very well be because they’re trans, nonbinary or disabled - perhaps for conscious reasons or perhaps because of other factors. The survey shows that there are more questions to be asked and that it would be worth paying attention to the experiences of these groups.