Reviews for Advances in Artificial Intelligence

The average rating for Advances in Artificial Intelligence based on 2 reviews is 5 stars.

Review # 1 was written on 2019-07-29 00:00:00 Jacques Hobscheid Where are we going? And will "we" still be here when we get there? Digital Soul is about the nature of our world when machines become as intelligent as humans and beyond. It is also about the nature of those machines. It is clear that Georges has thought long and hard about the subject, has read widely and has compared notes with other futurists. His expression is reasoned and reasonable. There are no muddy sentences or mystical ambiguities. He has worked hard to make sure that his ideas are accessible to a wide range of people including those with no expertise in the field of Artificial Intelligence. Clearly the problem is to derive benefit from super intelligent machines without letting them take over our lives. Georges believes that it will be difficult to do that since, as the machines get smarter and smarter and we allow them more and more latitude and we more and more depend on them, they will come to control us. But this is where I think Georges goes astray. The question I would ask is, would they WANT to control us? I think it is better to say that they will inadvertently control us or that we will condition ourselves to be controlled. Georges implies that human-like values, such as that of self-preservation will automatically follow from machines becoming intelligent. But actually the machines will have no values at all and no desire, either. They will have no inclination to act except as such inclinations are built into their make-up. Georges also implies that he knows what qualities or values are desirable in a machine. He speaks of "nicer, testosterone-free, superhuman beings" as opposed to "greedy, violent, barbaric, self-absorbed" beings. (p. 212) While these are surely agreeable preferences, it is not clear that artificial creatures designed according to human choice would long survive. It is also not clear that we would want to design machines according to human values. We would want to design them as tools (which they are) to assist us in following our desires and supporting our values. Notice the difference. Machines that work toward fulfilling the desires and upholding the values of human beings are not the same as machines that contain the desires and values of human beings. What I think Georges temporarily forgets is that no machine is going to "want" to do anything unless "desire" is built into the machine. The machine doesn't care whether it is plugged in or not unless we somehow encode such a desire into the machine. What Georges seems to assume is that somehow the complexity that we will demand from machines will somehow necessitate that we inculcate desire, self-preservation and the like into the machine. I think this will not be necessary at all. Indeed I suspect our machines will tell us that they will be able to function just fine without the institution of some kind of supercode or primary instruction telling them to protect themselves and have ulterior motives. (Such notions led to HAL 9000's murderous behavior in Stanley Kubrick's classic film 2001: A Space Odyssey.) I think a more likely future (and one that Georges addresses) is a symbiosis between people and intelligent machines in which the machines have the knowledge, skill and intelligence necessary for making decisions, but that the actual decisions and the impetus for action remain with human beings. However, should intelligent machines, as Georges fears, somehow acquire purpose and goals and desires such as self-preservation, then there is a great danger of our lives being taken over and controlled by intelligent machines. He warns us that we have to guard against that danger. Georges rightly brings up the Fermi Paradox in Chapter 18. Since it would appear (to some at least) that the universe is teeming with intelligent life, Fermi famously asked, "Where is everybody?" One of the many answers (aside from "we are alone") is that "technological civilizations have a very short life expectancy, because they promptly destroy themselves during their technological adolescence." This insight from Georges on page 214 is another way of pointing to what he is worried about. Still another way (perhaps) of expressing this is to say that we will merge with our intelligent machines, and having acquired a sort of superintelligence, will find that the values that were built into us by the evolutionary mechanism are muted, values such as self-preservation, curiosity, greed, anger, vengeance, etc. Any sort of desire may be culturally evolved out of us. Why do anything at all? may very well become the unanswerable question. Perhaps this is what happens to technological civilizations in their adolescence, and that is why we haven't heard from them. Beyond this I think we need to realize that evolutionary creatures, which we are, are just a place along the way to something else. What that something else will be is as much beyond our ken as understanding quantum mechanics is to bubble bees. Regardless of some disagreements this is a very interesting book well worth reading from cover to cover. I agree with his enthusiasm about artificial intelligence and I agree that we should continue to pursue its development and not become neo-Luddites. But I am not afraid of a future without human beings as we are now constituted. We are imperfect creatures. We are appropriate and adapted to the present environment. When the environment changes, as it surely will, we may no longer be able to adapt and may go the way of the dodo. So be it. We know from looking at the past that all species eventually die. New ones come into existence. Should the future be any different? As we see the limitations of humanity, as we see ourselves for the first time as we really are, perhaps it is time for a greater identification. Instead of identifying exclusively with human beings, might we not identify with a larger process that encompasses all life forms including those to come? --Dennis Littrell, author of "The World Is Not as We Think It Is"

Review # 2 was written on 2013-12-02 00:00:00 Gaw Masu Una buena introducción a los problemas relacionados con la concepción mecánica de la mente. Siendo una introducción, abre a los problemas sin reducirse a lo más superficial. Lo mejor, la bajísima carga ideológica y la exquisita honestidad con la que está escrito.