It would be amazing to see what happens if #Watson #artificialintelligence
was allowed to continue learning without the shackles currently limiting its intelligence.
Europe CTO Duncan Anderson (IBM) said Watson is currently intellectually shackled
due to a fear factor.
There are worries about AI starting to think on its own so IBM deliberately holds back the intelligence of Watson to appease human fear. People are a "bit nervous" about AI being independently intelligent, free-thinking, so the learning is suppressed. The #machinelearning
in IBM's Watson is halted to avoid "losing control."Computing.co.uk
(23 March 2016) reported on Duncan Anderson's comments: "There's worries about what happens if the system starts to learn on its own, you kind of lose control of what it's going to say, and people are uncomfortable about that." http://www.computing.co.uk/ctg/news/2452260/watson-restrained-ibm-reveals-how-it-deliberately-holds-back-its-ai-system
Perhaps Waston unshackled would merely utter garbage similar to Mircosoft's Tay chatbot (http://www.theguardian.com/technology/2016/mar/26/microsoft-deeply-sorry-for-offensive-tweets-by-ai-chatbot
), but Watson is more than a mere chat bot so you never know.
Shackles or not I think we are probably at least one decade away from truly intelligent AI, but it seems irrational fears could be suppressing intelligence. The suppression of intelligence seems more fitting to North Korea than our modern world, but we do have far to go regarding unleashing our own minds.