Is Artificial Intelligence Here?

 Artificial Intelligence  

Soon computers will be able to design themselves. Richard Waters of the Financial Times recently joined SciTech Now host Hari Sreenivasan to discuss the ramifications of truly intelligent computers.

Artificial intelligence has been heading in one direction for 60 years according to Financial Times West Coast Editor, Richard Waters.  And that direction is for the technology becoming, "as smart as people."

In this respect, the history of AI has been one of disappointment, as the early practitioners were way ahead of themselves in terms of the capabilities they could achieve. This was especially pronounced in the so-called AI Winter of the 80's and 90's, when it seemed that the HAL 9000 system of the movies was never going to be achieved.

"How will they explain to us, with our smaller brains, what they are doing to improve themselves?"

Now, the technology has caught up, and the field of artificial intelligence, including deep learning, is in a boom time. "Its not as though computers are as smart as people yet, but they've started to do things, particularly in the last two-to-three years that no-one thought was possible," says Waters.

Related articles Jason Silva Looks At Artificial Intelligence  The Rise of Artificial Intelligence  James Miller Discusses Aspects of the Singularity

Waters points to IBM's Watson defeating human competitors at Jeopardy! as one example, but beyond gaming, he points to the power of machine learning and Big Data to find answers that humans have a hard time finding on our own.  Medicine is one key example. Drug discovery and unraveling the complex interactions of biology are an area where AI might help out.

Commenting on the threat of artificial intelligence, Waters says, "How will they explain to us, with our smaller brains, what they are doing to improve themselves?" thinking about the potential for recursive self-improving computer systems.

Sreenivasan asks how this is going to impact society and democracy as super-intelligent machines begin to have more of an impact (in a time span of five-to-ten years, according to Waters).

SOURCE  SciTech Now

By 33rd Square

Embed