Sunday 25 February 2018

AI will be worse - but it will happen anyway... (Artificial Intelligence is the new bureaucracy)

As always, public discourse focuses us upon the wrong issues. With Artificial Intelligence the focus is upon whether it can be 'better' than the human mind - but the real question is much simpler: whether AI can replace the human mind.

The point is that AI systems can replace human thinking whether they are better or worse. Better or worse doesn't matter...

AI will, in fact, be worse than the human mind: much, much worse - but that doesn't stop AI from replacing the human mind.

AI is an extension of bureaucracy - which has replaced the individual human with systems of committees; has replaced judgement with votes; has destroyed responsibility.

Is a committee better than the individual - No, it is worse.

Has that fact prevented committees replacing individuals in all positions of significant authority in all societies in the developed world? No. Bureaucracy is everywhere, all the bureaucracies are linked, the individual is is a slave of The System - not the slave of a person; and this slavery applies to all individuals, even/ especially those who are members of the committees...

AI is bureaucracy cubed - it is driven by the same intent, it has the same objectives, and it has the same indifference to consequence. Just as bureaucracy is intrinsically immoral, intrinsically evil, in its destruction of responsibility - so exactly is AI - but this is no accident. On the contrary this is precisely why committees run everything; and why bureaucracy can (and will, if plans go through) be replaced by AI. 

AI is bureaucracy 2.0.  The question is, who - or what - stands behind the bureaucracy? What purpose drives this long-term agenda to replace all human thinking - that is, to replace all human thinking, by all individual humans, ultimately.

(Clue: the answer isn't human.)

All through our world, the worse-bureaucracy has replaced the better-individual; responisbility and judgment have been annihilated on multiple fronts... The process continues, globally, without evaluation or negative feedback -  and this is the exact intention for rolling-out AI.

Bureaucracy is always worse, but it has happened anyway. AI will be worse, but it will happen anyway. 

(Unless, of course, we stop it. And the place to stop it is in our own minds, our own hearts, our own deepest understanding. AI is a vampire - it can only enter where it is invited; but at present we are inviting-in the vampire - we are indeed paying the vampires to assimilate our souls.)


Unknown said...

Yes, this is one of the many dangers of "AI" as currently formulated. The problem isn't that "AI" will have true consciousness, hyperintelligence, and a soul. The problem is that "AI" as deployed will lack one or more of these things and then be handed the keys to the civilization. Not Artificial Intelligence, but rather Genuine Stupidity will destroy us.

The "AI" designers do not believe in Free Will, spirits, much less God or Jesus. I expect the results of their efforts will be ... poor.

-- Robert Brockman

Wm Jas Tychonievich said...

It's a good point that voting (i.e., decisions made "automatically" by an algorithm which takes various individuals' opinions as input) is a step in the direction of AI.

Bruce Charlton said...

@Wm - On the other side; voting has always had a (fake) moral imperative behind its introduction and extension - while AI instead has a (fake) notion of technocratic effectiveness.

Adil said...

For sure, we need to go back to rule by individual authority and place intuition before rules/bureaucracy. So if I may ask a question of political nature, what do you think about the argument that democracy is better att keeping checks and balances than authoritarianism where the people are subjected to the "arbitrary" decisions of a monarchical top head?

To me it seems that while some rulers/kings can be bought, in a democracy everything is for sale (at least in its modern form) - wheras a king can always be persuaded by justice. He has no time for micro-management and can only succeed by trust and competence in the end - and so everyone becomes more free.

At least from what I can see from your line of reasoning, the final solution implies localization of power and an end to parliamentary "big government"?