Thursday, February 17, 2011

Artificial Intelligence

Many people think of AI as 'failed' or being science-fiction. The thruth is, AI is everywhere now. Narrow machine-intelligences have become ubquitous in every part of life, ranging from medical diagnosis-systems over rocket-designing to jeopardy-winning supercomputers. All those technologies are the work of AI-research. Constructing a system that can make informed decisions based on an input is a task we humans have mastered over the last 20 years. The next step would be to design an AGI, an Artificial General Intelligence, an entity (as it is no mere system anymore) that has an own view of the whole world (or at least substantial parts of it) and can solve problems on practically every level of our reality.

Ray Kurzweil estimates the time for this to happen to be around 2029. He derived this from years of studying the exponetial curve that describes our growth in things like computational power. In this year the computational capacity of the human brain will be availible for 1000$ and software development will be sure to come up with something that uses this power in a most efficient way. By then our scanning of the human brain will be so accurate, that we can practically reverse-engineer it and have a model of that we know it will work (because it works for us). There are many ways through which we can achieve AGI and one will sooner or later succeed.

Once the human-level AI will appear, it will change everything. Literally, everything. Creating another type of intelligence, free from our human limits, is the single most greatest feat we will ever have brought forward. In practically no time, our problems, each hundreds of years old and fiercly complex, will simply disappear, because they are understood and can be solved. No human would ever have to be lonely anymore, because everyone will have an AI-companion, poorness and scarcity will be things of the past because we finally can manage our resources without too much waste. Ultimately we ourselves will 'upgrade' us and implement AI-systems in our own biology, eventually leaving it behind altogether, because it is extremely slow compared to electronical or optical systems. We will be able to 'reprogram' our thoughts at will, look anyway we want, simulate everything in virtual and 'real' reality. This time will be the great liberation of thought from it's mortal imperfect substrate. Many will still put this far in the future, but I say prepare yourself, The Singularity is Near.

All hail intelligence!

11 comments:

  1. I know kurzweil, From www.kurzweilai.net or something :D
    I'd really love to see the first ''AGI'' come alive!

    ReplyDelete
  2. hehe, just make sure humans always get the final word

    ReplyDelete
  3. we are slowly becoming more and more alien like. with AI robots, our lives would be easier but somewhat lazier. I see true potential within a AI community robot network for big operations.

    ReplyDelete
  4. Many think it will require Quantum computers to achieve human-like intelligence. In some ways machines are already smarter than us, think memory and maths.

    ReplyDelete
  5. I saw a documentary about a guy who installed a robot that was learning just like a baby thats crazy. I believe that something like 'Irobot' will happen in the future :P

    ReplyDelete
  6. AI is scaring me... who knows, maybe in a few years we will have androids

    ReplyDelete
  7. I sure hope Ray Kurzweil is right. I definitely want the singularity to happen during my lifetime.

    ReplyDelete
  8. I know that Jeopardy match with the computer scared the hell out of alot of people..

    ReplyDelete
  9. I can't wait till we get enslaved by machines

    ReplyDelete
  10. it's very easy to change the mind of machines!!

    ReplyDelete