Could AI become the real (almost) perfect politician?

Shall, perhaps, AI become the perfect solution for all political issues?

As the political scene becomes more and more complex and society, apparently, tends to do all its best to self destruct, the approach of the AI singularity could stand a solution for this matter. I should say in the best case scenario at least (whereas the AI will not wipe out the entire humanity from the surface of our planet).

I am reading more and more articles and topics about Machine Learning and how the computing power and logical intelligence is stepping forward. Operations that took days and days of work performed by a single person using a computer in only 15 minutes (cfr. "The wonderful and terrifying implications of computers that can learn") or computers that are going to predict crime or serial crimes yet about to happen (like the all Italian project "KeyCrime" which - for the record - is already working and bringing results). The possibilities and drawbacks of the Machine Learning industry are fascinating and promising.

As the technology goes forward, the political scene becomes more and more rough. Political issues are overwhelming the current planning capacity of our Politicians. I believe this is a world-wide problem, but I'll stick to what I can touch or see closely. Our today politicians are unable to take decisions. The human factor of trying to gain their own objectives shall be one of the main reasons. Gaining self advantages tends to be a priority over the whole society (see note 1 below). Add the massive amount of variables that characterizes today's reality and you'll have a chess table with millions of "pawns" and hundreds of millions of fields.

So could possibly Machine Learning and, suddenly, AI help us in taking better decisions? Most probably the lack of self interest (as in political and of self-interest like earnings) of a machine would solve most of the problems about the political issues, whereas it would elaborate a solution based on events and with best results. It would also be able to really calculate a solution even with a massive amount of data, and take a decision upon it. You may say that the "voice" of the single voter would take place in the equation. From a technological and technocracy point of view these things are amazing.

It wouldn't be moral or ethical

I do not agree. Morality and ethics can be "teached" to a system. How would you teach someone about ethics or morality? By giving inputs. Child abuse, killing, hurting someone, robbery. Those are values that can be relatively easy taught to someone, and therefore to a machine. If you take a crowd, the morality of that crowd would be the result of the values of each single individual. So, back to the above, the system should possibly act and reflect the morality of the community.

It would hold no spiritual values

That would be true, but it is not a politician's job to talk about spirituality, unless you live in Vatican City or a country where the political strength is held by Church (or any other religious movement). As a matter of fact, I won't get into the topic.

People are different

One of the strongholds of Communism was that everyone is equal, meaning there were no difference between people, albeit there are. Democracy, on the other hand, gives people the power to decide of their very life, but the reality is that each one's freedom ends where someone's else one starts. It wouldn't mean that different views shall not be taken in account. An "Artificial Intelligence based politician" (if we may call it that way) would take count about all the visions in one place. No internal conflict about what one party says or the other. Decisions wouldn't be taken because of behavior or dislike of someone's opinion.

Say, for example, that among 60 million people divided between two major political factions, most of both factions would need social help. One faction might be contrary to the decision, because of budgets, because they don't consider it right within their own vision. The other faction could not reach the majority to get the vote, and therefore it would result in a prolonged argument between the two factions and a massive expense of public budged. But the problem wouldn't be solved.
The true neutral politician (the "machine" you may say) would independently evaluate the problem, project the best solution following the moral and ethical principles it was taught with, and give an answer. In a matter of minutes probably. What could be possibly wrong in such a system?

Conclusion

As the technology is evolving at an increasing speed, so is society. The "millennials" will reach the age of vote by two years, and their way of living and thinking is slightly different than ours. So how could possibly a politician from the 80's do what is the best for them? As they are our bequest to this world, we can't tame them into ways of living and thinking that were completely different than now. It would require a change of view and elections by every year (and also, having politicians that have no experience in decision making). So, whenever we want that or not, a change in politics shall occur sooner or later, and machine based politics appears to me as one of the best solutions available. Perhaps inevitable, unless we want the entire system to collapse.

The human factor shall never disappear from the political scene, but it would radically change its role, since it would also be like playing chess on the board above against someone (actually "something") that is able to predict hundreds or thousands of moves while you are barely able to see the whole table.

Note 1: I should really correct myself here. Many (if not most) politicians are really capable of planning and working for their own objectives. Which probably overwhelms their capacity of doing any good for their electors.