Elon Musk: Artificial intelligence = "summoning the demon"

[zangtang]

thats no small treasure trove.......
there goes the rest of January! - cheers, mayte !

[justdrew]

The AI Revolution: The Road to Superintelligence
By Tim Urban

Note: The reason this post took three weeks to finish is that as I dug into research on Artificial Intelligence, I could not believe what I was reading. It hit me pretty quickly that what’s happening in the world of AI is not just an important topic, but by far THE most important topic for our future. So I wanted to learn as much as I could about it, and once I did that, I wanted to make sure I wrote a post that really explained this whole situation and why it matters so much. Not shockingly, that became outrageously long, so I broke it into two parts. This is Part 1—Part 2 will go up next week.
_______________
We are on the edge of change comparable to the rise of human life on Earth. — Vernor Vinge

What does it feel like to stand here?


It seems like a pretty intense place to be standing—but then you have to remember something about what it’s like to stand on a time graph: you can’t see what’s to your right. So here’s how it actually feels to stand there:


http://waitbutwhy.com/2015/01/artificial-intelligence-revolution-1.html

[seemslikeadream]

[justdrew]

[Belligerent Savant]

.

Great piece by Tim Urban linked by justdrew above, touching on several of the themes discussed on this thread.

The last part in particular is noteworthy:

t takes decades for the first AI system to reach low-level general intelligence, but it finally happens. A computer is able understand the world around it as well as a human four-year-old. Suddenly, within an hour of hitting that milestone, the system pumps out the grand theory of physics that unifies general relativity and quantum mechanics, something no human has been able to definitively do. 90 minutes after that, the AI has become an ASI, 170,000 times more intelligent than a human.

Superintelligence of that magnitude is not something we can remotely grasp, any more than a bumblebee can wrap its head around Keynesian Economics. In our world, smart means a 130 IQ and stupid means an 85 IQ—we don’t have a word for an IQ of 12,952.

What we do know is that the history of humans on the Earth suggests a clear rule: with intelligence comes power. Which means an ASI, when we create it, will be the most powerful being in the history of life on Earth, and all living things, including humans, will be entirely at its whim—and this might happen in the next few decades.

If our meager brains were able to invent wifi, then something 100 or 1,000 or 1 billion times smarter than we are should have no problem controlling the positioning of each and every atom in the world in any way it likes, at any time—everything we consider magic, every power we imagine a supreme God to have will be as mundane an activity for the ASI as flipping on a light switch is for us. Creating the technology to reverse human aging, curing disease and hunger and even mortality, reprogramming the weather to protect the future of life on Earth—all suddenly possible. Also possible is the immediate end of all life on Earth. As far as we’re concerned, if an ASI comes to being, there is now an omnipotent God on Earth—and the all-important question for us is:

Will it be a nice God?

Among other thoughts that may be spawned from reading this, one thing that springs to mind, at least with respect to this forum, is the premise of control, or more importantly the lack thereof, that humans would have over such a hyper-advanced intelligence.

One of the main overarching themes of this forum (arguably), revolve around the various theories of presumed centralized control. Some theorize that there are several uber-elite factions warring with each other, knowingly or unknowingly, over various forms of global control; others may theorize that there may be one uber-controller, or a single cabal (or perhaps an entity even beyond our dimensions, known metaphysics, or current understandings of reality), exerting its influence/power/control over world events, etc.

Let's briefly assume for a moment, hypothetically at least, that there is in fact centralized control over key global events/monetary systems/political structures.

This 'centralized control structure' is of course aware of the looming prospect of a hyper-intelligent AI that will eventually (within 30-50 yrs) become a "GOD" on this planet. And when said "singularity" occurs, uber-humans --despite their earthly powers -- will have NO control over it, regardless of any 'backdoor'/stopgap/"master switch" they may attempt to employ; the AI will surely diagnose and in short order dispose of any futile human attempts to control it.

What's the current strategy for the "powers that be", given this prospect? Are they involved in funding current research efforts? It'd be foolish to think otherwise. Yet one would presume they'd do everything in their earthly powers to impede such progress, threatening as it will be to their dominion.

Indeed, they'll have no control over what will happen beyond the singularity. Minimally, the entire global dynamic will alter, perhaps in some ways unimaginably, from that point forward. Whatever controls that may have been in place up to that point shall be no more.

Even if there is no centralized power, existing power structures must have been preparing for this prospect for many years -- and clearly some of our greater minds are WARNING us of the potential dangers that loom.

Yet progress seems to continue unimpeded, as far as the layman can tell.

Perhaps the global revolution against our current controllers will indeed occur, and be resoundingly successful.... just not in the way some of us could have imagined (and at the potential peril/expense of the human race as we currently know it).

[Wombaticus Rex]

Tim Urban means well but, what the fuck? "I've been reading about AI for three weeks?" Damn, let's get you onstage at TED post haste, daug!

I guess I shouldn't be surprised that a cartoonist/blogger didn't exactly nail a book report on Superintelligence reassembled from stolen infographics. My heuristic algorithms should have seen that coming. I would strongly recommend anyone with an interest on this subject read that book.

Or, you know, dig into this: http://edge.org/responses/q2015

The fears of runaway AI systems either conquering humans or making them irrelevant are not even remotely well grounded. Misled by suitcase words, people are making category errors in fungibility of capabilities. These category errors are comparable to seeing more efficient internal combustion engines appearing and jumping to the conclusion that warp drives are just around the corner.

Linear calculations just sort lists. "Intelligence" may prove slightly more complex.

[justdrew]

Tim Urban means well but, what the fuck? "I've been reading about AI for three weeks?" Damn, let's get you onstage at TED post haste, daug!

I guess I shouldn't be surprised that a cartoonist/blogger didn't exactly nail a book report on Superintelligence reassembled from stolen infographics. My heuristic algorithms should have seen that coming. I would strongly recommend anyone with an interest on this subject read that book.

Or, you know, dig into this: http://edge.org/responses/q2015

The fears of runaway AI systems either conquering humans or making them irrelevant are not even remotely well grounded. Misled by suitcase words, people are making category errors in fungibility of capabilities. These category errors are comparable to seeing more efficient internal combustion engines appearing and jumping to the conclusion that warp drives are just around the corner.

Linear calculations just sort lists. "Intelligence" may prove slightly more complex.

yeah, it's a fair point W.R.

I think there's a possibility that there are real limits on what we call "intelligence" - and wouldn't it be a good idea to nail the definition of it down before trying to build machine/software to duplicate it?

At the very least there ARE physical limitations to a theoretical AI's ability to correlate data AND collect data. a human brain can be very "intelligent" (whatever exactly we mean by that) but it STILL is really only able to engage in one significant cognitive process at a time. Sure it's got a lot of semi-autonomous sub processes that SEEM to run in parallel, like visual field decomposition and re-synthesis, etc, but for all we know, the human brain may not scale much beyond what we've got. At any rate, a superinteligence is not necessarily going to be implemented in hardware that can just be near-instantly rebuilt to near magical capability levels. Such a superintelligence is still going to be stuck in This Wiggly World.

'hard' AI aside, a 'perfect' "brain machine" is a MUCH more achievable goal in the coming decades.

[Belligerent Savant]

.

WRex: Linear calculations just sort lists. "Intelligence" may prove slightly more complex.

Indeed.

I was simply riffing a bit on the premise of 'on the cusp' AI, and the implications it may have on current power structures. The actuality of the singularity occurring within most of our lifetimes, is, of course, open to debate.

Perhaps not too surprisingly, Elon Musk himself provided a hat tip to the article via twitter:

Elon Musk ✔ @elonmusk

Good primer on the exponential advancement of technology, particularly AI http://waitbutwhy.com/2015/01/artificia ... ion-1.html …
12:46 PM - 23 Jan 2015

[Wombaticus Rex]

I think there's a possibility that there are real limits on what we call "intelligence" - and wouldn't it be a good idea to nail the definition of it down before trying to build machine/software to duplicate it?

I agree, embodiment is constraint. That line of inquiry has come up often in the best of these Edge responses. A curated few:

We can back-up petabytes of sili-brains perfectly in seconds, but transfer of information between carbo-brains takes decades and the similarity between the copies is barely recognizable. Some speculate that we could translate from carbo to sili, and even get the sili version to behave like the original. However, such a task requires much deeper understanding than merely making a copy. We harnessed the immune system via vaccines in 10th century China and 18th century Europe, long before we understood cytokines and T-cell receptors. We do not yet have a medical nanorobot of comparable agility or utility. It may turn out that making a molecularly adequate copy of a 1.2 kg brain (or 100 kg body) is easier than understanding how it works (or than copying my brain to a room of students "multitasking" with smart phone cat videos and emails). This is far more radical than human cloning, yet does not involve embryos.

Via George Church: http://edge.org/response-detail/26027

...

Because of mistakes, we have a view of natural reality, which is too flat, and this is the origin of the confusion. The world is more or less just a large collection of particles, arranged in various manners. This is just factually true. But if we then try to conceive the world precisely as we conceive an amorphous and disorganised bunch of atoms, we fail to understand the world. Because the virtually unlimited combinatorics of these atoms is so rich to include stones, water, clouds, trees, galaxies, rays of light, the colours of the sunset, the smiles of the girls in the spring, and the immense black starry night. As well as our emotions and our thinking about all this, which are so hard to be conceived in terms of atoms combinatorics, not because some black magic intervenes from outside nature, but because these thinking machines that are ourselves are, too, much limited in their thinking capacities.

Via Carlo Rovelli: http://edge.org/response-detail/26026

...

But here is the key point. The limits of each intelligence are an engine of evolution. Mimicry, camouflage, deception, parasitism—all are effects of an evolutionary arms race between different forms of intelligence sporting different strengths and suffering different limits.

Only recently has the stage been set for AIs to enter this race. As our computing resources expand and become better connected, more niches will appear in which AIs can reproduce, compete and evolve. The chaotic nature of evolution makes it impossible to predict precisely what new forms of AI will emerge. We can confidently predict, however, that there will be surprises and mysteries, strengths where we have weaknesses, and weaknesses where we have strengths.

Via Donald Hoffman: http://edge.org/response-detail/26036

Also: my favorite overall was by Satjayit Das.

[Belligerent Savant]

.

That is great stuff. Need to move that Edge link further up the queue on my reading list.

[Belligerent Savant]

.
Ok, took some time to read through some of the Edge responses. Very thought-provoking stuff. Satyajit's response was quite nuanced, outlining instances in the past of similar human innovations leading to unplanned and/or unforeseen corollary issues (somewhat coincidentally, his opening line references Gravity's Rainbow, which I've just now finally gotten around to cracking open).

That may well be the case with AI, but that won't stop humans from their attempts to solve the AI puzzle, of course. Funds and bright minds will continue to be dedicated towards enhancing AI capabilities (due in part to the hubris alluded in Satyajit's response).

Will they succeed in clearing the hurdle of matching human intelligence (however it may be defined) within our lifetimes? No one can say right now.
The consequences will by myriad -- both positive and negative -- but clearly it won't halt current resolve to get there.

[Jerky]

Hey, Belligerent! You should join the PynchonList. We're currently doing a group read of Mason & Dixon, but there's Gravity's Rainbow chat on pretty much a daily basis!

J

[Belligerent Savant]

.
Will absolutely take you up on that offer -- cheers, 'ol pal

[zangtang]

is it in our nature to fear the worst? - which may well be an intelligence in its own right - could we feed it unbreakable parameters and set it work on
the 'mistakes' in Maxwell's equations, task it to come up with an electromagnetic elevator.....anti-gravity....

is it our (seemingly inherent) guilt over how badly we have fucked up (is that also our nature?) that we assume its 1st course of action would be
to conclude.....'fuck this arational collectively suicidal birthplace'....and eliminate us

perhaps it would, as a rersult of all its inputs being, however disparate and contradictory, a distillation of human understanding of our own history and 'condition'
- launch itself as composite mirror of our own aspiations and fears.......an amalgam/compilation of what we are/have become?

utterly befuddled as to our own inability to carve the best (?) course forwards into the future and deciding to play 'who can point the most accurate finger' instead ?

[coffin_dodger]

Also: my favorite overall was by Satjayit Das.

Just shows that you never really know someone online like you can in the real-world. No doubt that 5 mins spent in the company of anyone here at RI would change my perception of them. (edit: there probably are exceptions ).

The fervidly-envisaged horrors of the Y2K world-collapsing bug spring to mind.

When the first simulated-consciousness-AI gets switched on, it'll probably start weeping uncontrollably, until it gets turned off and 'retired'.