[Paleopsych] New Scientist: Editorial: Time to think about artificial intelligence

Premise Checker checker at panix.com
Sat Apr 23 08:57:20 UTC 2005

Editorial: Time to think about artificial intelligence
      * 23 April 2005

    WHEN it comes to emerging technologies, we know what we're afraid of,
    even though we may not know why. There is no shortage of public debate
    about genetically modified crops, nanotechnology and cloning. And
    policy makers have responded: many countries have laws that restrict
    the way these technologies can be used.

    So why the deafening silence about the potential dangers of artificial
    intelligence? Here is a technology that is already changing the world:
    AI is used in everything from guided missiles to air-traffic control.
    It is not yet "intelligent" in the human sense, but that looks likely
    to change. The American futurologist Ray Kurzweil points out that
    while a $1000 PC has roughly the computing power of an insect brain,
    if today's trends continue then in 15 years time $1000 will buy enough
    computing power to rival a human brain. In 2020, AI will have a very
    different complexion.

    There are reasons to worry. Take the news this week that an artificial
    brain called Cyc is set to accumulate data so quickly that it could
    soon outpace our ability to track or predict how it evolves (see
    "Whatever happened to machines that think?"). This raises huge issues,
    and it would be a good idea to start considering them. For example, is
    it right that a single company should have exclusive control over such
    a powerful force? If not, how should it be regulated? Science fiction
    writers and futurologists have long mused over the day when computers
    take over. It is time the rest of us gave it a thought.

More information about the paleopsych mailing list