Will machines soon be smarter than their makers?

This is what researchers believe at the "Singularity Conference: Artificial Intelligence and the Future of Humanity", which took place this week. The problem: how do we prepare in advance for a situation where these machines will be hostile

In the center of a black hole there is a point known as the "singularity point", where the laws of physics cease to make sense. Similarly, argued the futurists who spoke at the weekend conference, information technology is rapidly moving toward the point where machines will be smarter than their creators. When that happens, it will change their meaning for the human race in ways that are hard to predict, they say.
The "Singularity Conference: Artificial Intelligence and the Future of Humanity", held yesterday and the day before in San Francisco, brought together hundreds of scientists, science fiction enthusiasts and high-tech workers from Silicon Valley. They gathered together and tried to imagine a future where computers would program themselves and brain implants could allow people to think at the speeds of modern microprocessors.
Artificial intelligence researchers have warned that now is the time to formulate ethical guidelines to ensure that these advances do more good than harm.
"We and our world will not need ourselves anymore," Rodney Brooks, professor of robotics at the Massachusetts Institute of Technology (MIT), told the crowd. On the computers of the future, he claims that "the difference between us and them will become a completely different question."
Eliezer Yodkovski, co-founder of Jewish origin at the Singularity Institute for Artificial Intelligence Research, which organized the conference, focuses his research on the development of what he calls "friendly artificial intelligence." His biggest fear, he says, is that some brilliant inventor will create a self-improving artificial intelligence machine that will be immoral and therefore hostile.
The first use of the term "singularity" to describe such a fundamental transformation in technology is by Vernor Wenge, a mathematician from California and author of the Bible. High-tech entrepreneur Ray Kurzweil promoted the singularity concept in his 2005 book, "The Singularity is Near", in which he claims that the exponential rate of technological progress makes the creation of artificial intelligence (smarter than humans) - the only logical outcome in the future.

Science or science fiction?
Kurzweil, the director of the Singularity Institute, is so confident about his prediction of the singularity that he even set a deadline for it: 2029. Most "singularists" believe that they have enough evidence to support their claims, citing as evidence the dramatic advances in computing that have taken place in the last 50 years. In 1965, Intel co-founder Gordon Moore correctly predicted that the number of transistors on a chip would double every two years ("Moore's Law"). For comparison, the supporters of the singularity point out that the stage of evolution of the human species from the primates amounts to a difference of only three times brain capacity.
With advances in biotechnology and information technology, they say, there is no scientific reason why human thinking should not be accelerated to speeds of up to a million times. But some critics mock the singularists and their obsession with "technological salvation", or alternatively "technological holocaust". These predictions are sometimes dismissed as more science fiction than science.
But supporters of the idea claim that ignoring it is irresponsible and could have serious consequences. "According to the forecast, we will eventually reach a point where we will create artificial intelligence," confirms Yudkovski. "The mere fact that it is impossible to predict exactly what will happen, and the exact date, is not an excuse for us to close our eyes and stop thinking about it."

To the 2007 Singularity Conference website