Yes, there are many aspects of what I suggest that are, in themselves, likely to cause wide spread change. Obviously developing the data systems and algorithms are a challenge to prevent bias. That is why transparency is critical. The great problems to date AI/ML decision augmentation is that the algorithms and data sets are proprietary and secret. We would need to build a public develop and system admin agency at the cabinet level to manage this if the US remains together. There could be no corruption by private capitalist ventures.
That by itself would change the nature of government. Within this the implementation of very strong EU model privacy laws could be used to breakup the technical monopolies.
I would visualize this working best at the urban region or small state level. That would allow the elimination, perhaps, of the federal legislative level. All decisions would be run from government level closest to the people. Have to think about that.