Wow. Mind blown. Thanks for this angle that I had not considered!

I think there's an increasingly compelling argument to be made at this moment in history, with the rise of powerful increasingly intelligent software that can 1) rigorously shape policy to ensure that ends are met and that 2) the means to those ends employ minimal bias, and 3) do so adaptively, shifting policy dynamically as circumstances evolve. I believe it's time to start planning for the day when we shutter the voting booths and find a better way to govern using networked sensors and dispassionate computers.

Democracy is dying. Long live automatic government for the people.

Is this really necessary? IMHO, yes. It's become clear that the easy manipulation of a gullible, clueless, and intellectually lazy electorate has broken all three branches of US government:

1) we have an imbecile President and a cabinet of grotesques who border on insanity

2) we have a congress that for over 30 years has done almost nothing constructive, that campaigns on one set of principles then serves others, the primary two being: the financial interests of Those With Money, and paying lip service to the fantasy-laden desiderata of Those Without.

3) a legal system that imprisons over 90% of those charged without a trial.

What brought America to such a dysfunctional State? The democratic popular vote, of course. Nothing else could.

Thus I believe it's time there was a serious discussion (outside of Washington, of course) about alternative ways to shrink and de-bias all three branches of government, at all levels.

I'm convinced that software soon will be powerful enough to fairly, effectively, and adaptively guide the players (both citizens and businesses) in questions of policy to reliably and precisely achieve desired ends, rather than relying on the venerably bent tools of power -- a gigantic incoherent mass of self-serving, ends-oblivious, and inconsistent principles chisled into marble long ago by power brokers to preserve eon-old biases that reliably serves only vested interests.

If we want AI to reach its true potential, I propose better government be its ultimate objective. Start from the bottom up -- local government -- and introduce ends-driven smart services to serve civic and main street business needs at that level -- where feedback is clearest, most immediate, and most accountable. This way we might just put government right, from the bottom up, cutting away Caesar as we go.

The idea that AI would rise up and seize power over its human creators has always seemed preposterous to me. But it is not at all preposterous, as your comment demonstrates, that we might imagine AI as some kind of savior and put it in that position intentionally. This is the real danger.

It is a prescription for dystopia. Either the machines, acting without understanding, will botch things up royally, or unscrupulous people will find a way to use the machines as cover while they pull the strings to their own benefit. Or both.

Yea... technology at large scale has, countless times, been a large detrimental to large amounts, if not all, members of societies. I can already see the hyperintelligent AI realizing the only way to force pliability amongst very slow-to-change socialized humans is what many throughout our history have also learned - force and slavery allow the cogs to adapt to where you need them. I'd rather not be totured because I don't modify my behaviors to keep up with the ever-changing, cost-optimizing algorithms that be.

Also are you suggesting that, instead of the standard dystopian model of AI overthrowing humans, that we, instead, have a revolution to put AI in the place to do all that it may? That would be ironic. More realistically, no one would join in the bloodiest revolution in human history for the glorious resolution of running our government on AWS.

Fictitious fun thoughts aside, and to quote J. Tillman, I'd rather not have the "world [be] written in lines of code." And, to be honest, your response is exactly the type of extreme SV character response I'd write in mocking hyperbole on a site where those very ideals might, quite frighteningly, actually be lurking.

I agree with your analysis of the problem. However, I deeply distrust your solution.

Two problems: One, do we really get unbiased software? Or do we get software that's claimed to be unbiased, but has actually been quietly biased according to the authors' prejudices (and all attempts to criticize it get shot down because "it's really unbiased, so the problem is with you")? Or do we really get a good-faith attempt to write unbiased software, that winds up biased in ways and for reasons that we don't understand?

Second, how are you going to make this happen? It's going to involve taking power away from the existing powers-that-be. And you don't sound like you're going to do it by democratic means, since you see democracy as broken. How do you propose to do it? ("Bottom up" isn't enough of a plan. At each step, it's going to be a war against entrenched interests. How are you going to win?)

I'm not worried that much about software bias - I feel that a lot of good could be achieved with software clear enough for all sides to accept it represents their interest in an ok-ish way.

What I'm worried about is this: technology has always been a force multiplier for those who wield it. For now it's us (maybe in the future it will be a superintelligent AI, maybe not). Even if we manage to write a pretty unbiased governing software, how are we going to ensure that it isn't subtly manipulated by some interested party? And can we ensure that while also making the system extensible, so that it can cope with the demands caused by the impact of ever-progressing technology on our societies? It seems to me that the two goals - a system above humans and a system that can evolve with us - are fundamentally incompatible.

Democratic popular vote did not lead to a dysfunctional state. The lack of democratic vote has led to the problem. We have two entrenched parties because of our winner take all system, gerrymandering designed to take away voter rights, laws restricting rights to vote, and vast media conglomerates.

Back at the dawn of the US, the one with the most votes became President; the next became Vice President. This was a good idea, as it would at the very least encourage some bipartisan discussion. Perhaps the VP would be found more often on Capitol Hill, in the role of President of the Senate.