October 12, 2008
from EducateYourSelf Website
And it’s a future that they’re working
feverishly to make a reality today.
Google’s methodology in their quest is to exploit and harness the powers of emergence, while adding ‘parts’ that perform cognitive tasks in their own right. The idea is to push the term super-organism to the fullest potential. The insights are the ant colony, and the beehive.
The models are the Internet, and the
human brain. The entire premise of emergence is ‘the sum is greater than
Each human mind is a hive, of neurons; a hivemind.
Each neuron is abysmally stupid, and each brain
part is nearly worthless alone, yet the interconnected cells of each part
and each part interconnected give rise to consciousness and higher brain
But what if a collective of intelligent beings
harnessing global scale supercomputers armed with state of the art
algorithms made the goal of turning their symbiotic co-evolving component of
the Internet their life’s work?
It’s the company slogan, and when you hear them speak it they place much emphasis on “all”. But what’s more, according to both co-founders of the company, they intend for it to “understand” what all of the worlds information means. In many cases, they’re going to extremes in order to “gather all” of it, but they’re lettings us do the work for them wherever they can.
A look over their “Product” list is as far as one must go to get an idea, yet it goes further than that.
But first, the main search feature must be
This includes content from social networks,
blogs, news sites, and so forth. Feeding a link into
Archive.org’s Wayback Machine paints a
candid picture of this process. Their urgency to archive all possible
acquirable data also extends to other areas, such as your personal life
via Gmail, Health, Calendar & Google Desktop (which scans all of the
files in your “personal” computer) .
Add to that
Docs and Knol.
But note their new “Android” mobile device service, which has 2 alarming features.
So on one hand it conditions you to submit your
eye iris scans, and on the other it conditions you to embrace constant
real-time GPS tracking of your every move. The latter is dually striking as
virtually all modern cell phones already embody GPS tracking, except most
people aren’t yet aware of that.
For some insight into the implications of their machine eventually having in its ‘intellectual possession’ virtually the entire history of humanity’s books, newspaper, scholarly academic papers and so forth, consider the statement from a Google Factory Tour guide:
The point is proven in the fact that they’re
scanning the entire books whether or not the entire contents will be
browsable online. Perhaps my publishing of this article online is giving the
Machine even more focused insight into itself?
Much of their epic archiving quest wont even
immediately pay off, but it’s being kept as fruit waiting to ripen, or
rather waiting for their conscious entity to ripen to be able to harness it.
By this point many would declare that AGI isn’t possible, but regardless of
beliefs and possibilities (external possibilities aren’t dependent on one
persons beliefs), the stated goal exists.
But the Googler’s seem to be right on course to
reach their ambitions by conscious direction of that delves beyond mere
emergence alone. There are many cases of the 2 Google co-founders going on
the record about “AI”, and between them and their related media’s, it’s
quite clear that their intentions aren’t merely ‘narrow AI’ nor is any of it
Begin with Google Video and shortly afterwards they acquired Youtube. These are both sites where the user does the work in providing the profitable content for them. For some it was neat for Youtube results to appear in Google Video searches, but then Google began crawling most of all other streaming video hosts including many of their competitors.
During that shuffle, Google acquired Neven
Vision, the worlds most advanced machine vision firm. One desirable
prospect for Google was that NV’s technology was already geared for mobile
devices. Another was that it was designed for both still photos and videos.
It can be understood as advanced biometrics that’s designed to recognize all
types of objects, not just human faces.
But I see a scarier side: When their infantile systems grow conscious enough, not only would they - or It - know and understand everything humanity has ever written, it would also have to a certain degree all of our videos from film to personal cams.
One side is it helps it become sentient, the
other is it accelerates its ability to understand humans individually and
socially. After all, you couldn’t expect a machine to become conscious &
intelligent without vision, nor could it understand humans without
seeing them in action.
Meanwhile, their Translate efforts has
their system rapidly learning how to translate any language from any
language, guided by a handful of engineers who in most cases don’t even know
the languages themselves. You can see this by doing the typical search, and
you can bet they’re already working on integrating the technology into audio
In May of 2008, Google hosted their own “Machine Learning Summit“, of which,
Prior to that, in 2006, internal documents
leaked stating their plans the build “the worlds largest AI laboratory”.
That lab might be already existent somewhere in their own properties, or it
might be in or set to be in a government / military facility.
Being the size of ‘2 football fields’, it sounds
reminiscent of what everyone used to say in reference to the NSA’s
system that was and is used to monitor virtually every form of
telecommunications in the US and much of the Earth.
This would especially be the case if you had literally a million or more parallel platformed CPU’s at your disposal (like Google does). Consider that computing power per $1000 is literally less than millionth what was during the Manhattan Project, and that project only cost about US$24 billion.
Anything even resembling a modern semiconductor
computer hadn’t even been invented yet. Meanwhile, every year their
capabilities expand as CPU prices drop and work gets easier, exponentially,
Moore’s Law and the
Law of Accelerating Returns.
Since the US military maintains global supremacy via its Navy, one could almost Google to roll out plans for ‘naval’ data-centers (which they have). The full scope of government & military involvement with Google’s AGI project goes well beyond the scope of this analysis, but suffice to say that Google have in 2008 signed a 40-90 year lease, with their geographical neighbors, at NASA’s Ames Research Center (ARC), in Silicon Valley, for a 1.2 million square feet collaborative research facility.
But not only does Google get to build on the
government land there, they already have exclusive access to land and park
their private Google jumbo jets on ARC’s “Moffitt Field”.
It’s also been alleged by a former CIA agent that not only did the CIA fund them during their earlier years, but that the CIA has an actual office in the main Googleplex headquarters, while it’s a fact that Google hardware runs the US Intelligence Community’s ‘spy wikipedia‘. DARPA claims the fame of inventing the Internet, and by visiting their website you can browse through their extensive list of various inter-related “cognitive” “self-aware” artificial intelligence projects.
Vint Cerf, Google’s “Chief Internet
Evangelist” VP, ‘invented the Internet’ together with Bob Kahn via
DARPA. Vint still works with NASA on the “Interplanetary Internet”, as well
as on other projects for the United States military.
NASA, in recent years, has pursued the perfect program to mesh with this. The goal of the “Intelligent Archives” sounds familiar to everything Google is doing in terms of “understanding” massive amounts of data. They even used the phrases “self-aware” and “cognitive”. The projects webpage now speak in the past tense, and it’s unclear its true status after its parent division became the “Intelligent Systems Division”.
In any case, with deep integration with NASA and
their Earth sciences, and the military and their hundreds of top secret
satellites, and AI all in-between, I’d say Page has found the perfect
scenario to pursue this dream.
By using Google Search we can tell the system is very much in perpetual tune with its own complete data-holdings.
One could argue that they’re flirting with
“hard-takeoff” AGI emergence on this front alone. And if you spend enough
time putting complex worded search strings ‘into’
Google Calculator and Code Search you might
help make that a reality.
That might make people reconsider that Google
fought for the new wireless Internet spectrum, and are working to bring 3
billion new people the Internet, good things.