Monday, July 17, 2017

The growing "menace" of AI

"I fully expect that in five years, due to Moore's Law, Cloud Computing and good old Yankee ingenuity, our home AI will be able to perform over 100,000 tasks. Most everything will be voice activated, so you might as well throw your remotes and your QWERTY keyboards away. Yes, it probably will be that good."

Hold on to your laptops or tablets kids. Elon Musk is getting everyone all lathered up once again about the growing "menace" of Artificial Intelligence (AI). So much so, his main motivation for going to Mars is to escape Earth. Escape? From what? I know that Steven Hawking said we needed to "get out of Dodge" in the next 100 years before we become par boiled due to global warming, but what the what is Musk talking about? He is talking about living through a real life episode of I, Robot

Really? Is AI really going to be a menace to our society? I thought AI was going to be a blessing. Mr. Musk said that in his business, he has studied the field of AI quite deeply. And what he found was quite disturbing (to him at least). Like, does AI learn? Absolutely it learns. How much so? One of the more complicated games on Earth is the Asian game of Go. Not that long ago (just a few months in fact), a computer program which housed an AI Go program wiped the floor with the reigning (human) champion. Kind of like what IBM's Big Blue did to the "unbeatable" Garry Kasparov in chess in 1997.

Here is the nut of the nut as far as Elon Musk is concerned. AI learns differently than humans do. And because AI learns differently, it thinks differently. That being said, when AI evolves in everyday practice, it might relate to us more like aliens than humans. And now we are back to a bad science fiction movie.

Musk thinks we need the heavy hand of government to keep this thing from getting out of hand. Good luck with that one Elon! Remember? We live in America, where technical innovation usually wins the day. And once a company patents an advance in AI, that company is going to be very reticent about sharing it with anyone - especially the government which (as we know recently), can leak like a sieve. 

I, like many of us, have a rudimentary AI device in the house. It happens to be a new Echo (Amazon Prime Day purchase), recently upgraded from our Alexa Dot. Alexa has recently gone from being able to perform a few hundred tasks to about a thousand. And because Alexa made the biggest and earliest splash on the market, it is now king of the hill. But hold on - this is America. Google has seen the success of Alexa, and decided to not only see Alexa, but raise the stakes on Alexa. When the new Google Home is out in the market, it will be able to do 10,000 tasks. Ten times what Alexa does. See how this is going?

I fully expect that in five years, due to Moore's Law, Cloud Computing, and good old Yankee ingenuity, our home AI will be able to perform over 100,000 tasks. Most everything will be voice activated, so you might get ready to toss your remotes and your QWERTY keyboards. Yes, it might be that good.

But hold on Bird - what about the "menace" of AI? I guess I am not too worried about it. I believe the marketplace will continue to develop AI to advance the common good of mankind. Besides - we have something AI will never have. We can always pull the plug.

1 comment:

  1. This topic could keep me awake at night. Thankfully, I am closer to the end of life than the middle and the dangerous and I think accurate predictions regarding this technology is a couple of decades away.
    While human-machine cooperation is a hopeful avenue to explore in the short to medium term, it is not clear how successful this will be, and by itself it is not an adequate solution to the social issues that AI automation poses.
    These constitute a major crisis of public policy. To address this crisis effectively requires that scientifically literate government planners work together with computer scientists and technologists in industry to alleviate the devastating effects of rapid technological change on the economy.
    The cohesion of the social order depends upon an intelligent discussion of the nature of this change, and the implementation of rational policies to maximize its general social benefit.
    I am not sure that governments are ready, nor capable of managing these processes. Nor defining social benefit. This could be "Snowflake" to the extreme limits.
    But having the intelligence to balance our fiscal realities with what is needed for social benefit may be the help our government needs to continue its existence. Or it may dramatically change and improve the way government interfaces with our lives.
    Scares the hell out of me, glad I won't be around to see the results, good or bad. Cause my money is on Gates, Musk and Hawking.