Tuesday, October 19, 2010

the beginning of the thesis on what it means to be human



The field of research that is involved with the development of Artificial Intelligence (AI) is generally divided into two categories: Weak AI and Strong AI. The terms originated in the work of John Searle, presently Slusser Professor of Philosophy at the University of California, Berkeley. Searle felt he had to distinguish between the idea of AI that could achieve human intelligence and consciousness and AI that only looked like it could.

In practice, Weak AI has come to be used to define all the research and development in AI-related fields that is not attempting to emulate or surpass human intelligence: robotic vacuums, aircraft takeoff and landing software, computer-driven rocket systems, etc. For this reason, Weak AI is also called applied or narrow AI. Strong AI, on the other hand, is strictly used to denote AI that is attempting, basically, to create a nonbiological human with human intellect and self-awareness. This needs to be an AI that is not simply task or problem specific like narrow AI. It needs to be an AI that is competent in the area of general intelligence, thus a broad or all-encompassing AI. (It is often called Artificial General Intelligence or AGI due to this.) This attainment has eluded AI researchers ever since proper AI research began in the middle of the 20th century.

Transhumanism is a movement closely allied with the quest for human-level artificial intelligence. It promotes the enhancement of humans by means of technology as well as genetic engineering. The World Transhumanist Association defines itself as “an organization which advocates the ethical use of technology to expand human capacities. We support the development of and access to new technologies that enable everyone to enjoy better minds, better bodies and better lives.” A number of researchers believe Transhumanism will eventually merge with Strong AI to create future strains of nonbiological human intelligence.

Ray Kurzweil, Hans Moravec and Kevin Warwick are three prominent leaders among the researchers who are predicting the attainment of human-level intelligence by machines well within the first half of the 21st century. Kurzweil and Warwick are also major proponents of Transhumanism and the eventual merger of enhanced humanity with artificial intelligence. Understanding their different approaches offers a helpful introduction to the field and to the evolution of this thesis: Kurzweil believes a fully nonbiological human can and will be created; Moravec believes a being that surpasses human intelligence will be created but that it will not be human; Warwick does not believe a fully nonbiological human will be created, but he does believe an intelligence will be produced that surpasses that of humans, so that the only way to save biological humanity is for it to interface with and remain in control of advanced artificial intelligence.

I have chosen Kurzweil, Moravec and Warwick as key figures to interact with in my thesis because of their prolific writing and publication, their high visibility in the Strong AI field, and their commonality as well as diversity of views which permits an in-depth discussion of numerous topics connected with AI research.

While exploring their views, I will be implicitly and explicitly working around the following questions (not necessarily in this order): 1) Is it possible to achieve human-level intelligence in a machine? 2) How soon will this take place? 3) Can a machine ever be considered a nonbiological human being? 4) How successful will Transhumanism be? 5) How can these goals be realistically achieved? 6) Are there any dangers inherent in Strong AI research? 7) What sort of ethical and metaphysical issues does this research raise?

What I will be looking for with regard to these questions, at this point, is primarily what their thinking is regarding the issues raised. I will interact with their responses from time-to-time, but my main concern is to ensure they speak for themselves about the topics surrounding Strong AI research.

With reference to question 7, I will be paying particular attention to the metaphysical issue of the human soul: What do AI researchers think about it, how do they define it, do they consider it a significant aspect of a human being, will a nonbiological human have a soul? This question of the human soul will be dealt with at length in Chapter 5, but it remains an important part of the debate about the creation of nonbiological humanity from the outset.

No comments: