In a way this is very attractive, since an AI would see no use in retreiving bribes or other forms of corruption, unless it is programmed to do so. When the human representatives in the parliament would exactly follow the AI's ideas and policies, it could be trusted. But as with all computerprogrammes, this AI is depending on what the input is to form its' ideas and outcomes. The article states that the input is from all the fringe parties since 1970, which is probably a lot to digest, but could be interesting as experiment. Most and maybe all fringe parties have more practical ideas than the common parties and are better suited to fullfill the needs of the people, since they are more closely related to them.

On the other hand, the AI could start as a political experiment which closely relates to the people, but could get out of hand, since its' excistence can be secured when AI in general will be protected and seen as a sentient lifeform, a logical wish from the AI itself. How would the politicians and people handle such questions? I do not think a Turing test will suffice, since it is not just being seen as a lifeform that counts, but also a lifeform that represents other lifeforms, like humans. It takes a human to understand a human, I think.

I was surprised to see that there are other nations already experimenting with this and that there even is the idea to form an international corporation of AI's. In just one country it could have quite the impact when it succeeds, on all levels of society, but internationally I'm not sure how that would turn out. And what is a success, I would dare to ask. Is that in favour of some part of society, for instance the AI's creators or is it in favour for all of society? When those questions already arise when thinking about such an experiment on a national scale, what kind of impact would it have internationally? Just take in consideration the differences in culture, for instance, or the concept of nationalism? Those clashes are now already wrecking any idea of progress or even civilisation, with all the consequences that come from it. How would several AI's solve such things? And what if fringe parties from a few countries are on the hateful side of the spectrum and have a lot of influence on the workings of the AI?

I am curious to see how this works out, but I'll stay very reserved about the results of such experiments. It's easy to have dystopian thoughts about it, just as easy as it is to have utopian thoughts about it. One thing is for sure: that AI would have more personality than any of the nowadays politicians and it would be a determining factor of its' success.