A governance computer would reflect the views and behaviors of whoever programmed it...
 
Well people reflect the views and behaviors of whoever created or educated them! Similarly a large team of programmers working together could conceivably, over time, write a really thorough set of judgment programs. It's a bit like if a really big group of people adopted an orphan and decided to train it to be the best Senator, President, or assistant Agricultural Secretary the world has ever seen. They teach the orphan civics, they teach the orphan mathematics and probability, they teach the orphan social compassion, ethics, thriftiness, and everything that they can think of to help it do it's job. At the end when that orphan turns the proper age to become what it was guided to be, if the orphan is intelligent and diligent it is going to make a pretty excellent officiate, especially if the orphan has had a long period of internship.

The only difference in substituting a computational system for the orphan (outside the technical complexity which we are coming to address quickly) is that the mechanical orphan can have a thousand people teaching it ideas at once, it never needs to sleep, and perhaps most importantly: it will not need to retire after a few short terms of service as many of our elderly officials do now, before any significant progress can be made. It can continuously be updated and improved forever with open-source code transparency so that anyone can go look up how it's mind works and the justification for it's decision-making... which we almost never truly see with out human politicians today. Then we can focus on developing all of the craziest economic forms like actual post-scarcity, which has an entirely separate group of opposed dissidents.