AI could potentially help democracy by being a less biased expert, something people could turn to if they don’t trust what is coming out of human experts. An AI could theoretically consume complex legislation, models, prediction market information and data and provide a easily questioned agent that could present graphs and visualisations. This could help voters make more informed choices about important topics.
Can this be done safely and reliably? If so can it help people make better decisions around AI safety?
[Question] Is working on AI to help democracy a good idea?
AI could potentially help democracy by being a less biased expert, something people could turn to if they don’t trust what is coming out of human experts. An AI could theoretically consume complex legislation, models, prediction market information and data and provide a easily questioned agent that could present graphs and visualisations. This could help voters make more informed choices about important topics.
Can this be done safely and reliably? If so can it help people make better decisions around AI safety?
Is anyone working on this idea currently?