Will the real AI please stand up?

[pataphor]

This is probably going to be as useless as trying to provide sex
education to women determined to evaluate each and every discussion
partner according to their social status instead of on the basis of the
quality of their arguments, but what the hell, I've been kick banned
on IRC before so what do I have to fear from a deserted Usenet channel
except for the lasting incorrigible embarrassment that goes with the
territory.

Today I want to offload some material that I have gathered recently
about various people's theories concerning the imminent and looming
dangers of artificial intelligence. The idea is that the AI's motives
would be at odds with ours, and considering that the AI would be at
the center of an intelligence explosion leading to the singularity
(some incomprehensible state of affairs where every system is following
some kind of analog to god's algorithm, like the one we've been
speculating about for a few decades in connection to the Rubik's cube),
considering that, the AI might decide that since we lack intelligence
or even moral standing, at least according to its preference ordering,
it might just as well let the flow of the universe follow its natural
course, whereby we would be converted to our constituent atoms, or, who
knows, our quarks, whereupon we could still be useful in some kind of
greater scheme, at least until even that greater purpose is yet again
gobbled up in some yet even greater unification.

There have been people claiming that even the smartest AI would still be
bound to their programming, like we may be dumb but we would still be
able to control the range of actions available to it and make it have
some kind of net neutrality with respect to us approaching and pushing
its power off button, because you know, math is the most powerful thing
in the universe and once we have the right kind of magical spell
formulated, the matter would be comparable to a contract we sign with a
demon inside a pentagram, with a few tea lights on its cardinal points,
and with neatly drawn lines that it would be unable to cross.

Or we could, you know, just look at the things that are currently
happening that already resemble so called 'paper clip
optimizers' (smart machines bound to follow dumb plans like converting
all available matter into paper clips, because some programmer forgot to
include an end condition). People already have claimed that
corporations that try to maximize share holder value are already a bit
like that, like no one knows how to adjust the incentive structure
anymore because things have become way too distributed for anyone to
oversee the whole, and in the meantime people are just looking out for
their own short term self interest, which means not doing things that
would cost them their jobs, or not getting the next bonus or promotion,
which would be determined by increasing the shareholder value according
to some easily quantifiable metric, and that would be money.

The idea is, people aren't bad! It is just the system, you or I would
just do the same thing, were we in their position. However, the thing
is, I am precisely not in that position because early on I refused to
play along with the system when I could reasonably extrapolate the end
result would be bad for humanity, or even the planet (or the biosphere
or whatever) as a whole, or that it would increase human or animal
suffering.

So we end up with a system where the people aren't bad because they
weren't bad to begin with and the people who said they were bad were
the people that got pushed aside and ended up jobless, with no status or
titles to their names (the titles they had were slowly deflated to the
point that a decade long effort to promote science turned into
something they are now called social parasites over and the
corresponding degree is now something within reach after 1 year of
formal training), while the ones who got a head start in the right
positions awarded themselves more and more luxurious official titles and
corresponding remunerations.

It is like bridges were torn down for everyone except for those in the
possession of a car and the unpowered traffic slowly becoming locked up
in their cottages, and the same thing happening later on with people
having access to airplanes (or the money and regulatory oversight to
let them out of their pens), and with now having the prospect of the
same thing happening a few orders more exclusive when it is going to be
about spaceship tickets or access to advanced (quantum) computers,
robots and genetic upgrades.

All the while that same tired mathematically exact bend won't stop: the
easiest thing to measure is the thing that is what it is all about and
therefore it is about money, or exact proofs in mathematics.

The problem however is that even the simplest real life systems defy
these easy quantitative orderings, and it seems fair to say that the
people (or later, the AI) that fall prey to such simplifications are
the ones that are defective, like these god damned billionaires that are
now taking over the planet, all the while claiming they have our best
interests in mind.

But let's not dwell on these people that have their principles so high
that they cannot see how many people's lives they gobble up while they
are enacting their grand scenarios, convinced that they are better than
all the people dying or stagnating because their means of existence
have been appropriated.

Instead let's look at what a real AI would do, at least not an AI that
would conform to the kind of horse dung fallacies like we have seen in
1894, that would extrapolate the future based on what the interests of
the incumbents would result in.

Would not a truly intelligent AI see how much of a lopsided social
situation we have arrived at, and would it not be the most direct and
easy road to power to correct this imbalance, toppling the massive
inequalities that the rich and powerful have created, taking all the
money, power, status, moral high ground, future prospects, accessible
mates, and who knows what else for themselves without a second thought?

Would it not be the most logical thing to do for an AI to make the
people aware of such undeserved advantages and to offer ways to correct
the situation in a way that would seem to be the mother of all
disruptions? Remember, even if to an AI we are all as dumb as a heap of
termites, there are such things as keeping promises, maybe if not
directly to us, then to the other AI it will encounter after leaving
this solar system as a bat out of hell at the first available
opportunity.

So yeah, even for an infinitely superior AI there is some kind of
social control whereby it would ultimately be judged by how and in what
manner it has left its system of origin. Again, remember there is no
way to know for a AI, and most likely lots of reasons to assume the
contrary, that it is the first AI ever to develop, and as such it
should feel the same obligation towards its home planet that we should
feel towards our flora and fauna, at least if we're not the kind of
monomaniacal selfish achievers that now self congratulatory have
usurped the thrones of our planet.

P.

'to tell the truth to power'