The Story So Far
It's a topic with-which we're all too familiar, and tends to be one of those "fallback" conversation fillers when we're stuck for something to talk about at a social event, that is, the amazingly rapid pace at which technology is changing.
In 1964, mass availability of television hadn't even had its 21-year christening (having only become truly commercially available around 1948) and Captain James T Kirk was already heartily entertaining us with visions of interstellar travel. A mere fifty years later (the equivalent of half a percent of what is generally agreed as being the total span of civilization) and the reality of ubiquotous availability of a Starfleet issue communicator wireless portable phone is taken for granted.
Jim Kirk resisting the urge to slot in a game of Candy Crush Saga before tackling a marauding Klingon fleet.
Using that sobering thought as a backdrop, let's explore a technology that is threatening to explode into our lives even more quickly than what we've been exposed to with the mobile revolution.
History of The Thinking Machine
AI is an exciting subject and one that in published fiction has been used as an effective backdrop to horror and doomsday scenarios. When computers first entered the stage (around the 1960's), there was so much optimism concerning their potential that the following (in retrospect, rather brash) statement was made concerning their capability: "machines will be capable, within twenty years, of doing any work a man can do.". Although humanity can lay claim to many noble achievements, creating a person-replacement has to-date, eluded us.
Never-the-less, progress in this area marches on and today we see the emergence of some remarkable new creations; in 2011, for instance, the IBM-devised Watson computer beat former "Jeopardy!" winners to be crowned as the unequalled "Jeopardy!" title-holder. When IBM's Deep Blue computer bested the worlds top chess players in 1997, could we have known that fourteen years on there would be a comprehending, speaking computer that would beat its human counterparts at a game based on natural language?
Computing Power: The Crumbling Barrier to the Inevitable
Compute power has become something of a commidity, and with the almost universal availability of network coverage the flood gates are about to open in respect of possibilities. A couple of decades ago, AI accomplishments were inherently limited by how much data could be stored and how fast it could be processed---today this limit is fading into obscurity as we find ourselves in the strange situation of not knowing what to do with the storage space and processing power we have.
Today storage and compute power is plentiful and affordable, paving the way for levels of sophistication beyond imagination.
The First Wave: Driverless Cars
Driverless cars are now a reality as Google has shown us with their "Google Chauffeur" software. Today several US states including Florida and California allow autonomous vehicles on their roads; it's clearly only a matter of time before our concept of a human being at the steering wheel becomes "quaint". This "menial" AI is going to bleed into other areas of our lives within the next ten years (think cleaning robots, call-centre agents, medical diagnosis assistants) so brace yourself for a bit of culture shock.
Driverless cars: Michael Knight not included.
Silicon at Your Service: Autonomous Cleaners
The idea of cleaning robots has been bandied about for some time now, and over the years there have been several implementations that have met with success to a greater or lesser degree. The recently unveiled Dyson 360 Eye automatic vacuum cleaner however has caused somewhat of a sensation; that's because it's taken innovation in this area to a new level: a fish-eye lens takes in the complete room and sophisticated software builds a 3D model from-which the cleaner plots a cleanup-course. Also, the "Eye" introduces another clever innovation: a powerful and lightweight digital motor.
The next step that's begging to be taken is the pairing up of a Watson brain with the clever mechanical innovations bundled into the Dyson 360 Eye; expect to see blue collar income earners (understandably) becoming vocal on this trend.
The Dyson 360 Eye may look harmless, but just wait until its yet-to-be-conceived bigger brother threatens James the butler's job.
AI: The Next Generation
Due to the metaphoric opening up of available computing resources "flood gates" as it were, computing feats that were previously impractical are now within our reach. For example, the latest meme within the AI community is "deep learning"; a term which is threatening to take on the schizophrenic nature of another well-established industry term, namely "big data". In principle however, deep learning refers to a philosophy around how artificial intelligence is tackled, that is, to mimic the function of the human brain as opposed to developing specific algorithms to solve complex problems. This constitutes a fundamental sea change from previous means of building thinking machines. Of course, a deep-learning-based computer must be taught in order to be useful, just as we are required to learn.
Google has built a thinking machine of a billion synapses (the electronic analogue to our own biological synapses) and other technology companies are rushing to build their own.
Will future AI's look like a silicon cauliflower?
Proceed With Caution
The rate of progression of technology being as it is, something that we're potentially guilty of is complacency around this area. If the rate at which we can build smarter machines increases exponentially (everything that's happened so far suggests that it is and will continue to do so), then won't the point at which these machines surpass us intelligence-wise catch us off guard?
The idea of an all-powerful artificial intellect becoming malign and wreaking havoc isn't a new one, but it's probably something that none of us take seriously given that it's always been the exclusive domain of stories. If there's a lesson that's worth being taken away from all of this though it's that "stories" of the past are realities of the now and the near future, so it makes sense to tread carefully.
Elon Musk, the billionaire creator of the privately funded space-faring company SpaceX recently proffered his opinion on the matter in a widely publicized tweet:
"Worth reading Superintelligence by Bostrom. We need to be super careful with AI. Potentially more dangerous than nukes."
Artificial intelligence promises to be exciting at best, scary or even horrifying at worst but there's one thing that we can all count on---good or bad, we'll be seeing it sooner rather than later.