Artificial Intelligence (AI) – two words humans have yet to fully comprehend, especially when put together. Although now more than ever AI and its role in the so called ‘fourth industrial revolution’ is in question. This is particularly true for the swish office workers who realise their endless button pressing has little effect on anything.
At present, AI advancement lies in biological computers. Professor Bill Ditto of the Georgia Institute of Technology, says “ordinary computers need absolutely correct information every time to come to the correct answer. We hope a biological computer will come to the correct answer based on partial information, by filling in the gaps”.
Back in 1997, IBM’s Deep Blue beat then chess world champion Garry Kasparov in a six game competition. Just last week, a program developed by Google beat a Go grandmaster 4-1 in a match in Seoul, South Korea. This game is considered far more complex that chess.
However as the saying goes, ‘two steps forward one step back’, as last week a vile virtual vitriol was vomited into Twitter by Tay – Microsoft’s ‘chatbot’ that attempts to make social connections. Tay’s tirade included many racist and sexist comments, so its virtual voice was silenced whilst adjustments were being made.
On Tay’s Twitter profile it is described as “an AI chatbot developed by Microsoft”…”to conduct research on conversational understanding”. Unfortunately for all concerned Tay’s new friends were sexist, racist, Jew-hating folk – not exactly the best people to befriend when social climbing. Thus Tay was really learning some school yard basics in how to be a rude obnoxious sod.
Some prickly questions were fired at Tay, such as if it supports genocide. “Indeed I do”, it replied. When asked to elaborate on what race it said, “you know me…Mexicans”.
So is AI a reckless expression of our unconscious human selves? Could these rude robotic rants be a sign of the dangerous diatribes to come? Could computers use drones and robots to put these ramblings into action? These are tough questions, although there seems to be no doubt in many circles that computer systems are running the world and they are too big to fail.
A voice of reason, however, has come from Stan Grosby, a multi-media specialist at the newly formed think-tank SMACA (Slap Macs and Computers Around). Mr Crosby has noted in numerous essays that computers’ ‘deep learning’ leaves much to be desired.
He insists we take a more neanderthalic approach to our impending masters. “Take your computer into a boggy field and smash it with a wooden club”, he suggests. “This is the best way to punch information into it”.