As an innovative enterprise, Google have a preference for artificial intelligence has always been. Not only acquired over more than 10 related businesses at an early age, even a Google search itself, especially image search to make good progress in recent years, can also be seen as part of the artificial intelligence. But go for Google is still a challenge. Since 1997, deep blue beat the world chess champion human chess seems to go the only remaining stronghold. Although in the beginning of the last century, mathematicians proved that chess is cracked, but at present and in the future for a long time unable to rely on brute force brute force, but must look for other ways.
This exhibition is to promote Google and a great opportunity to go, but it is only part of the scientific experiment, any formal competitions will not be lost and move on to the next, there is no precedent for playing chess with each other don't know each other. Because scientific tests need black box, Li Shishi cannot be tried before the next few games or watch the moves of the game, so the contest was not fair and equitable. Li Shishi after losing in three sets, the fourth in his pants was found under a program Bug, there was applause and cheers! But the cheer was given too many other meanings, especially in the sense of man against the machine. A team against a team, no matter how the results are essentially is the people's victory, and human wisdom is still an unknown mystery.
Charm game these days is a two-way process: players on the first day of the machine very defiance of blind confidence, and subversive just need three or four days, quickly turn to pay homage to the machine, the fourth game ended when magic disappears, people have found that people can overcome "God". Whether it is morbid or later to start the shock, is essentially due to the machines (science and technology) not very understanding. After the man-machine war could deepen public understanding of chess and artificial intelligence, then this game even without white. In the current discussion, there is a saying that is sensational, it's AlphaGo's victory means that the coming singularity. However, all of singularity many fuzzy areas in question exists, and therefore below the singularity of the births, contents and sort out the mistakes one by one.
What is the singularity (Singularity)
Singularity the etymology of the word comes from the Latin, meaning a single lonely existence (a being alone), medieval English, means unusual behavior or a single target in 1893 and became mathematical terms, usually translated as singularities (the singularity), in 1965, while building the model of the universe using Einstein's general theory of relativity into physics, cosmology become a physical term. In the field of artificial intelligence, in fact, 1958 Ulam (Stanislaw Ulam, this decade hit Monte Carlo algorithm was founded by von Neumann and Ulam) referred to the singularity at the commemoration of the von Neumann, 1965 I.J. Goodall (I.J. Good) use the term "intelligence explosion" (Intelligence Explosion) to describe this concept.
Today, however, we may hear more science fiction writer Vernor Vinge coined (Vernor Vinge) or entrepreneur Ray Kurzweil (Ray Kurzweil, the representatives of the singularity is near) predictions, they believe the coming decades so that a singularity will come, when the machines will replace humans. And in the case of physicist Stephen Hawking, entrepreneur Bill Gates, Mr MASKEY (Elon Musk) and others against the artificial intelligence also implies a singularity exists and can destroy the human meaning.
On odd points itself of pros and cons defended online are has many, force is of like Tim Urban AI revolution (online spread of thanks Panda June of articles is translation of this paper), against of like Toby Walsh odd points may forever not coming, although in I seems which finished burst former, at least is serious of researchers published in pre copies of articles, and former completely is a hype of stunt. But still would like to emphasize is that this is not really a science question, but a scientific prediction, a history of the human sciences issues. Different from the natural sciences, in the field of Humanities, if not each other entirely logically refute even one counterexample a lot of lack of positive examples, this interpretation is still a possible explanation for this model.
Singularity discuss nature and rationality
Singularity is not a new problem in the 1970 of the 20th century, the prevalence of catastrophe theory, is an example of Cambrian explosion of biological evolution to explain the explosive development of science and technology, have now declined, but the analogy with biology and physics, for example, what is the difference between? Judging from the vocabulary, the technological revolution and technological singularity and no essential difference between the two, the latter at best just objects reduced to the field of artificial intelligence. So, why did the meaning of an expression has been used to earn eye of one another?
The industrial revolution and the social revolution has "revolutionary" concept with rotten, revolution is everywhere. Only once the early scientific revolution in the history of science and technology, capital of the scientific revolution (The Scientific Revolution) was the birth of modern science. This concept is extended to many fields, namely lowercase scientific revolutions (the scientific revolution), such as the scientific revolution Newtonian physics science and Einstein's scientific revolutions, and can segment down. However, the emphasis on the technological revolution is very important for Internet companies, for innovative companies like Google, not just the primary productive forces of science and technology, technological innovation is the lifeline. Schumpeter's innovation economy, the trend is more and more obvious, everyone can see examples of our App on your phone to remind you every few days to upgrade, brush your sense of place. The conceptual level, too, they have to constantly use new concepts for packaging and promoting yourself, 4.0 and Web2.0, industry and so on. Although did not any new content, but its commercial value and social significance, even smaller influence on the development of science and technology, not here. This itself is neither good nor bad, but for discussion but not limited to this.
In addition, since economists, economic history of Metrology introduced, measures in the field of the humanities to shine. But, not to mention the public does not know the measurement for the historical study of functions and limitations, many professional researchers know is that fitting out a beautiful model, and not to think about the essence behind it, its root cause lies in the excessive worship of numbers and mathematical models and have only a smattering knowledge that mire of philosophy of science and technology say technical reason.
In Singularity theory the most widespread view is, after the coming Singularity artificial intelligence is growing exponentially, human civilization is the linear growth or stagnation. However, this argument is not clear, it there is no indicator to indicate in which artificial intelligence with people, any one who has a clear understanding of the measurement law knows that, choose what kind of metrics as criteria for evaluation results will differ, because indexes are value load with researchers. Heerqi·kelafu (Helge Kragh) cautions in the introduction to the historiography of science Cliometrics only as an aid to historical research, it is necessary to measure very carefully. Still some people believe everything that we as a community can be calculated, which is the restoration of the old mechanistic, we can only say that some of the indicators for a reasonable fit to explain some phenomenon, but we may never know the nature of the phenomenon.
Take a step back and say, even if there are so-called singularity, exponential growth is not required. Exponential growth is discussed in order to strengthen their own views and have the choice to choose, not just exponential functions, and many other functions such as polynomial functions, logarithmic functions, mathematical Singularity exists said, but the pace of development is completely different. But different indexes such as e^10000t and e^0.000001t in the vicinity of 0 are within close proximity, but the singularity when the estimation error is several orders of magnitude. Scholars shouted in Qufu Confucian the hometown
Secondly, history and how to predict the future is a very confusing problem. Measurement method for history, most people used natural extension of fitting curve, assume that the future is bound to be like this, and ignores the reality of the development of conditions have been met. A clear example is that family planning nothing like absolute population growth curve before and after, and then by model fitting is a very bad explanation. No one can predict the future, history didn't have a law, what law is. What is the phenomenon? Measures suggest that the past one of the standard under development is based on the phenomenon of exponential growth is; what are the rules? no matter how the future will still index development this is the law. Today law or law (law) has and revolution as was abuse has, but real of science law has never not this forecast, science law of forecast sex actually is a general suitable sex, lift a not rigorous of example, today I in Besal oblique Tower threw two a size varies of two a iron ball while landing, I forecast tomorrow you in tour Tower threw three a different iron ball also must while landing this is law.
Confuse the natural law and the social sciences is most terrible, which is why no one can accurately predict the future, as in 1965, Simon (Herbert Simon) said that within 20 years the machine will replace humans, there is no doubt that any similar comments by his impact. Borrow Clark (Arthur Charles Clarke): when a distinguished but elderly scientist States that, what it must be like in the future, that he must be wrong. Even once in a while someone says the pre-sleep is just a myth, actually we backtrack and development must always find similar words, can find an appropriate set of indicators to fit only after we know, this is the nature of historical research. From this point of view, if there is a singular point at the moment, and only after that we really have to know.
Historians are also the forecasts of metrology to be merged is because in modern society, prophecy and foreknowledge is very important for the future, all social and economic activities are based on the forecast. Youwaer·helali (Yuval Noah Harari) elaborated the modern in the brief history of the human sciences and similar to the forecast for the future of the financial system and homology, as said above, if you choose a good measure, in real conditions under the premise of not changing, not fitting for a long time considered to be a good explanation.
Finally, the singularity and artificial intelligence discussions is actually two separate propositions. We think you will find that attitude towards Singularity Singularity proponents are not exactly the same, strange is this contradictory state of mind usually appear in the same person. For those Internet businesses, they need public support of the artificial intelligence on the one hand, presented a singularity, but also faced with another problem, how to eliminate everyone to the singularity replaced human fear of artificial intelligence. As Kevin Kelly (Kevin Kelly) said that technical elements and technological development without human control, said the singularity draws near, on the grounds of human singularity can do to absolve their own research, which is almost passing the buck.
Where are the starting point of singularity? People with artificial limbs not count? Mechanical heart for people not count? A man with an electronic brain is not, does a country with an electronic brain count as? Upload only, or to not only upload or download to the body? Is the world into a semi-mechanical or electronic only, first person even if it is the lab? Technology on the alienation of people actually started early, but people have no awareness. Do people still think glasses or cell phones very strange, and still others leave you? Why only now discovered because 400 years ago, Columbus "discovered" America, this is a typical phenomenon of folks. I think the singularity is a actually a fuss about vocabulary, the good news is that this alienation or human development more apparent its bad side to sharpen the antithesis between technology and people, like tomorrow I am not I, robot to seize power. In fact, you are not you, so-called embedded or in the future, virtual reality are nothing more than another form.
Therefore, we cannot to a singular point of uncertainty and ignore the changes taking place around us, in fact, those changes are affecting the development of artificial intelligence. No singularity, we still have the potential to develop Super Intelligent AI, singularity, does not mean that AI will replace humans. In fact, human reason needs to be a turning point in large part out of a sense of ceremony needs, but what happens to you is a stretch of consecutive, something a clear transition from quantitative change to qualitative change, some are not. Singular point may not be a good model, but it is still a useful model.
Conclusion
I have always felt that the most terrifying prophecies are not science-fiction fantasies of all kinds of human extinction, for example Skynet and the like. The contrary, it is because the fantasy scene, formed the imagined community, human scenes in accordance with their development, and finally died in his prophecy. This is a very old philosophical jokes, it makes me shudder, and the world of tomorrow movie reminds us that should not fall into obsession with predictions about the future. Strange of is, now many science fiction fans into this false of Carnival among, they on Science terms of familiar instead hinder has they on related problem of further thinking, a Skynet of joke open once two times is humor, old took with said thing I not see its Ming also, they while shouted with human Center doctrine undesirable, machine on people care, on the himself and into human building of joke among. Really good science fiction, whether painted superhero, AI, or aliens, are explored through his own, as Lei·buladeboli (Ray Bradbury), say, science fiction was never to predict the future, but to prevent future.
Now more and more carefully, is feeling more and more civilized. Like with all the assets of bet, a guy that only 100 blocks compared with a billionaire, which is clearly much more carefully than the former. From the human point of view, people should take responsibility for their own future, should consider the consequences of technology, indeed the social ethics, legal systems have not kept pace with the technology, such as automatic driving, the law provided that the drive must be a natural person, if not a change in the law, does the drive automatically after you want the driver sitting in the passenger seat to help monitor?
Today all on artificial intelligence of research and discussion, are no from von Neumann computer and brain of paradigm, mass and entrepreneurs of description also far said not Shang rigorous, entrepreneurs and scientists of role should is each other added of, former let everyone concern artificial intelligence, which let everyone know artificial intelligence what is what, currently seems this independent thinking and each other promote of scenes also far not enough to let people satisfaction. After deep blue beat chess champion, impact of artificial intelligence research, impact on human society, after all these years is still far from enough. Similarly, for AlphaGo technology, thorough discussion is far more important than the results, independent of the media and the public to think a long way to go.
No comments:
Post a Comment