Sunday, July 20, 2014

superintelligence


mashable |  Humans are currently the most intelligent beings on the planet — the result of a long history of evolutionary pressure and adaptation. But could we some day design and build machines that surpass the human intellect?

This is the concept of superintelligence, a growing area of research that aims to improve understanding of what such machines might be like, how they might come to exist, and what they would mean for humanity's future.

Oxford philosopher Nick Bostrom's recent book Superintelligence: Paths, Dangers, Strategies discusses a variety of technological paths that could reach superintelligent artificial intelligence (AI), from mathematical approaches to the digital emulation of human brain tissue.

And although it sounds like science fiction, a group of experts, including Stephen Hawking, wrote an article on the topic noting that "There is no physical law precluding particles from being organised in ways that perform even more advanced computations than the arrangements of particles in human brains."

Brain as computer 
The idea that the brain performs "computation" is widespread in cognitive science and AI since the brain deals in information, converting a pattern of input nerve signals to output nerve signals.

Another well-accepted theory is that physics is Turing-computable: That whatever goes on in a particular volume of space, including the space occupied by human brains could be simulated by a Turing machine, a kind of idealized information processor. Physical computers perform these same information-processing tasks, though they aren't yet at the level of Turing's hypothetical device.

These two ideas come together to give us the conclusion that intelligence itself is the result of physical computation. And, as Hawking and colleagues go on to argue, there is no reason to believe that the brain is the most intelligent possible computer.

In fact, the brain is limited by many factors, from its physical composition to its evolutionary past. Brains were not selected exclusively to be smart, but to generally maximize human reproductive fitness. Brains are not only tuned to the tasks of the hunter gatherer, but also designed to fit through the human birth canal; supercomputing clusters or data-centers have no such constraints.
Synthetic hardware has a number of advantages over the human brain both in speed and scale, but the software is what creates the intelligence. How could we possibly get smarter-than-human software?

10 comments:

BigDonOne said...

No problem---> http://www.japantimes.co.jp/news/2014/07/19/national/science-health/chimps-inherit-intelligence-prove-humans/#.U8wWCPldWSp

CNu said...

Don, you ignorant slut...., do chimpanzees have natural language and generative grammar? If not, then you are easily the stupidest and most unteachable anal fissure of ALL TIME! Think.about.it...,

BigDonOne said...

When gut bacteria won't fly, try language and grammar...??


Genes define IQ potential at conception - environment after that determines the extent to which that potential is reached. Humans, chimps and doggz.......

ken said...

I don't know Don, its probably not as silly as an idea as climate change and trying to make us believe one major factor is the majority of cause versus the actual reality of factors and possibilities, but its pretty close.


http://www.huffingtonpost.com/2013/03/25/china-engineering-genius-babies_n_2949121.html

"Hank Greely, director of Stanford’s Center for Law and the Biosciences, says preimplantation genetic screening could one day render procreation via sex obsolete, at least for those who can afford it. But that doesn’t mean it will result in a generation of geniuses. “I think it’s pretty clear that intelligence -- if it even exists as an entity, which remains controversial among psychologists -- involves a boatload of genes and genetic combinations, all of them substantially mediated through the environment. The chances that genetic selection is going to lead to really substantial increases in human intelligence in your lifetime are low.”

Munné agrees. “IQ is controlled by probably more than 1,000 genes, so there is no point even trying to control for that,” he says.

The problem is simple math, adds Lee Silver, a genetics expert and molecular biologist at Princeton. Even if you could pinpoint a handful of genes that were likely to result in a higher IQ, the chances of any given embryo containing the right combination are minuscule. “Add in the fact that nongenetic factors account for 40 to 50 percent of the variance of something like intelligence,” and the project is basically hopeless.

CNu said...

Chimps and dogs don't have discernable natural languages serving as a basis for the cultural acquisition and transmission of knowledge and information across generations simpleton. Your repeated demonstrations of limited comprehension suggests that your own mentality may fall somewhere closer to that of a dog or a chimp, but you shouldn't generalize from your own struggles to the rest of us language-users: In its final form, Binet's
test provided an index of scholastic performance based on the
prevailing standard of scholastic success. In other words, scores on his
test generally correlated with the ratings assigned by French teachers
in the classrooms of his day.
By using teachers judgements of classroom performance as the standard
by which his test was validated, Binet established a practical basis for
its use as a predictor of success in the school system. Because his aim
had been to identify children who required special schooling, he did
not require, nor did he assert, a theory or definition of intelligence.
Moreover, he did not make a distinction between acquired or congenital
feeblemindedness and he never argued that poor performance on his test was a sign of innate mental inferiority.
On the contrary, he sternly rebuked his contemporaries who contended
that intelligence is a fixed quantity that cannot be augmented. http://subrealism.blogspot.com/2007/11/ii-is-establishment-reviving-eugenics.html

woodensplinter said...

Evidence from human famines and animal studies suggests that starvation can affect the health of descendants of famished individuals. But how such an acquired trait might be transmitted from one generation to the next has not been clear. A new study, involving roundworms, shows that starvation induces specific changes in so-called small RNAs and that these changes are inherited through at least three consecutive generations, apparently without any DNA involvement. The study, conducted by Columbia University Medical Center (CUMC) researchers, offers intriguing new evidence that the biology of inheritance is more complicated than previously thought. The study was published in the July 10 online edition of the journal Cell.

The idea that acquired traits can be inherited dates back to Jean Baptiste Larmarck (1744–1829), who proposed that species evolve when individuals adapt to their environment and transmit the acquired traits to their offspring. According to Lamarckian inheritance, for example, giraffes developed elongated long necks as they stretched to feed on the leaves of high trees, an acquired advantage that was inherited by subsequent generations. In contrast, Charles Darwin (1809–82) later theorized that random mutations that offer an organism a competitive advantage drive a species’ evolution. In the case of the giraffe, individuals that happened to have slightly longer necks had a better chance of securing food and thus were able to have more offspring. The subsequent discovery of hereditary genetics supported Darwin’s theory, and Lamarck’s ideas faded into obscurity. http://newsroom.cumc.columbia.edu/blog/2014/07/17/effects-of-starvation-inherited/

Dale Asberry said...

Little Don, STFU: http://www.sciencedaily.com/releases/2010/05/100524143416.htm

BigDonOne said...

Bacteria, along with nutrition, education, and other parenting variables, are merely part of the environment associated with the degree to which the Hardwired Genetic Potential for IQ is achieved.

Easily proved by looking at extreme cases -- Identifiable genetic conditions such as Down's, William's, and severe myotonic dystrtophy all tend to produce cognitive deficiency. No amount of any kind of environment, other than perhaps a brain transplant, will ever boost those afflicted genetic losers into the right-half of the IQ curve:
http://en.wikipedia.org/wiki/Down_syndrome
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1049802/
http://ghr.nlm.nih.gov/condition/williams-syndrome
http://www.ncbi.nlm.nih.gov/pubmed/8071955

You cannot "fix" the lost cognition in above hard-wired genetic defects with any environmental therapies. Ditto for most of the left-half genetics in the general population. The Chinese realize this and are working on it to breed super-IQ's. http://newobserveronline.com/chinese-government-funds-1-5-billion-drive-to-discover-genetic-basis-of-iq/ They will ultimately eat our lunch......

CNu said...

Tell us what you know about this hardwired genetic potential for IQ you keep harping about? What evidence do you have for this hardwired potential BD? Here are the facts as I know them: http://subrealism.blogspot.com/2014/07/why-not-try-and-breed-kwisatz-haderach.html#comment-1495441406

BigDonOne said...

Smart kids tend to run in families. The genetics causing this are complex and need to be identified. But it won't happen in the USA because any researcher who goes there, even to speculate, is ostracized due to the un-PC nature of such. China will do it... http://newobserveronline.com/chinese-government-funds-1-5-billion-drive-to-discover-genetic-basis-of-iq/

The American system blindly squanders Billionz$$ on the bogus PC idea that "gaps can be closed" and "outcomes can be equalized"....

As for hard-wiring, that is merely the term BD uses to try and get thru to the less perceptive that one's genome, particularly that which defines brain development, is fixed at conception. You even acknowledge "the genetic component that contributes to IQ test scores is highly heritable (somewhere around .50)." Actually, when all the environmental noise is filtered out, the correlation will ultimately be around 0.999, just don't have the data yet.