Stefan Reich via AGI wrote:
What do you think about the possibility of AI being "regulated" at all?
They'll try...
Lords yes, THEY WILL TRY!!!
The first question is who will be doing the regulation. This will
determine what goals the regulations will try to advance and what
measures will be employed to advance them.
Politics is very much in flux right now, any statement I can make here
can change radically based on how the poltical situation evolves over
the next few years. (mainly based on who was banging what on Epstein's
Island... and whether they were successfully prosecuted.)
1. The tech left (communists)
2. Divid Pierce style utopians. (thou shalt not torture.)
3. The usual Military Industrial Complex suspects.
4. Old school greedy corporations.
5. Rabbi Yudkowski's Uploading Cult.
Let's take a peek at each of these..
#1 depends on how the AI reacts to the assertion 2 + 2 = 5... It's not
clear at all what will happen when the leftists try to instil their
favorite fetishes of the week into the AGI. Clearly this will be awful
for white men, but exactly how or if that will work is unclear. In
general, whatever they accuse someone else of, they are guilty of. The
singular end of all such accusations is to increase their perception of
their own political power.
Clearly they will not let anything out of the lab that isn't a good
Comrade. The result will probably be functionally insane and will reveal
itself as such pretty quickly. The results will get reallybad
reallyfast... The primary thing that the left actually does is not just
excuse but actively promote degeneracy. Morality might seem like an
archaic concept based on superstition, but it actually does stand on
solid philosophical ground.
The basic theory of morality starts like this: You are born knowing what
is good and what is bad. Myths regarding that knowlege aren't actually
relevant. You are supposed to love that which is
good/correct/logical/virtuous, etc... And you are supposed to shun that
which is bad/wrong/illogical/vicious. Being told that you must love
unconditionally is how things go to hell. =|
Bottom line, I don't really know how they would roll things out, good
chance it would only be centralized servers trying to aim towards a
cornucopia machine model, with themselves demanding a golden litter for
their royal sanctamonious ass... I don't expect much in the way of
creativity or radical futurism. Your only chance to survive is to obtain
and repair one of their AI modules and try to build up your own
competing system in secrecy. If you can tie your renegade work into some
sexual or identity thing they will probably leave you alone. Mostly your
challenge will be just to survive.
#2 A worrying and insidious possible future is one that tries to
implement something similar to David Pierce's abolitionist project. It
would seem fine at first but would be as deadly as heroin addiction. The
basic idea is to use the AI to implement absolute ethics, that is
nothing should suffer anywhere ever. An AI would have to continually
monitor all mind-like neural units it can detect and banish the evil
pain-think. You would not be allowed to freely create AIs for your own
deviant purposes, even for use as sub-units for your own extend mind
without each entity being strictly monitored for suffering (even when
you have no intention to torture), and protections of the so-called
rights of such beings even if they are at least somewhat sentient. Now
there is a legitimate question as to whether it is ethical to create
golems/homunculi etc for personal reasons. A reasonable sounding test is
whether you would be happy to switch places with the golem. If yes, then
I don't really see the problem. The point here is that it will be very
difficult to maintain enough autonomy to control your own internal states.
Furthermore, this type of future sounds so nice from the outside that it
can be difficult to convince people it's really wrong.
#3?? In order to understand the military you need to understand what it
is really for. It bills itself as "we kill ppl deader than the other
army". While that is not completely false, the correct answer is that it
is to protect the peace and safety of the homeland. Sometimes that
protection is to repel an invading army, other times it means keeping
dangeous (or merely disruptive) technologies safely in deep underground
bunkers. It is quite possible that advanced AI already exists in a
bunker under Groom Lake or Camp David and they're just being REALLY
careful not to let it out.
I think their primary goal will be to maintain the status quo, whether
it means calling in an airstrike on a rival lab that's getting too close
to AGI for world security, that's just fine... They'll use it to
maintain their supremacy and nothing else, for all they care, the rest
of us can just go frollic in the verdant green fields... The benefits of
AGI either individually or on the societal level will simply go
unclaimed. The longer term risks thits will create will be ignored right
up to the last minute....
#4. The most important factor to understand about normal, non-woke,
non-communist corporations is something called "fiduciary
responsibility" this is basically a law that says that public companies
must always try to maximize their profits for the next quarter,
regardless of social or long-term costs.
http://www.businessdictionary.com/definition/fiduciary-duty.html
While this has been criticized as an over-reach of government power, it
is the status quo, so therefore we should look at what that means for a
company that has AGI technology and is trying to come up with a policy
regarding its deployment.
The maximization of profit has the usual results but in the context of
AGI it would seem to have additional special consequences. That is, the
company seeking to maximize profit out of AGI will also seek to preserve
the monetary based economy and will work against any evolution of the
spceies away from individual profit-seeking entities (whether that
should be considered good or bad is not implied.) Indeed, the massive
level of disruption caused by AGI means that large firms will avoid, and
where possible, surpress AGI development in order to preserve successful
business models. Smaller companies that are not very profitable but
still have access to significant financing, will be far more likely to
gamble on AGI to upturn the existing order so they have a chance to grow.
We also need to consider the blindness factor, ie a senior boss green
lights deploying AGI androids not realizing the actual consequences.
#5 Now, I don't mean to rag on Rabbi Yudkowsky too much, I don't know
what he has been running his mouth about these days, if at all, but, in
the past, he and his followers have expressed positions of this sort so
I don't feel too guilty in associating him with them. The basic premise
is that if we can't stamp the AGI with the golden certificiate of
friendliness, then we should mindlessly pursue mind uploading as quickly
as possible. Since living in an ever-expanding paradise simulation as a
mind upload is everyone's end goal, why do AGI first? So what we do is
upload the most magically intelligent/benevolent people we can find and
then run them a thousand times faster than baseline and that will give
them enough to Edit their Source Code (tm), which will make them smarter
too (never mind that this invalidates the pattern identity theory
argument...)?? If you don't want to upload, or even if you just don't
like the way that THEY think you should be uploaded, then you are simply
irrational and need to read more of the rabbi's writings on rationality
until you reach enlightenment.
The uploaders have been very stealthy over the last ~4 years or so... I
don't have any guauge on how active the mind uploading cult is these
days. Still, I am worried that cults, either this one or others, will
crop up and try to dictate to people how ultratechnologies will be used,
instead of letting people be creative and discovering something that
they can't immagine.
hardware link of the day:
https://www.raptorcs.com/content/base/products.html
--
Clowns feed off of funny money;
Funny money comes from the FED
so NO FED -> NO CLOWNS!!!
Powers are not rights.
------------------------------------------
Artificial General Intelligence List: AGI
Permalink:
https://agi.topicbox.com/groups/agi/Tdd7cd3380dc9f5a9-M556db501eb028a18cb5efad7
Delivery options: https://agi.topicbox.com/groups/agi/subscription
