Cortana property of Microsoft Studios/343 Industries.
With advances in technology, electronics, and cybernetics (the electronic analog, not bionics), humanity is seeing amazing new inventions appear at a rapid pace. One of the principle focuses of several organizations such as Google (and even the CIA), is to build a quantum computer. For those who are unfamiliar, a quantum computer differs from a digital computer in that they use quantum elements (atoms and molecules) as elements for processing and memory, allowing unprecedented speed and flexibility. Rudimentary ones exist, but a functional has not yet been made. The important point to focus on, though, is what may result from the creation of a quantum computer. Namely: Artificial Intelligence.
Now, there are already things that we refer to as AI, such as the “intelligence” programmed into enemies and friendlies in video games. Different computing systems or pieces of software may leverage “AI,” but what they are really talking about is an adaptive piece of software that still operates within the constraints of its environment and is strictly limited to its programming. No new information can go in, and no creative output is possible. While these programs can sometimes be very convincing (as evidenced by attempts to spoof humans into chatting with pieces of code online), they are not really what scientists, inventors, and (most) science-fiction authors would consider artificial intelligence.
Now, there are already things that we refer to as AI, such as the “intelligence” programmed into enemies and friendlies in video games. Different computing systems or pieces of software may leverage “AI,” but what they are really talking about is an adaptive piece of software that still operates within the constraints of its environment and is strictly limited to its programming. No new information can go in, and no creative output is possible. While these programs can sometimes be very convincing (as evidenced by attempts to spoof humans into chatting with pieces of code online), they are not really what scientists, inventors, and (most) science-fiction authors would consider artificial intelligence.
When it comes to artificial intelligence, real artificial intelligence, many people, smart people, are terrified. And rightly so. To lend credence, Stephen Hawking and Elon Musk count themselves among those who are very concerned with the responsible development of AI, with Musk putting up millions of his own money to fund a think tank to research the subject. Now for those of us who have to work nine to five and don’t get to spend all our time speculating and inventing, this may all be a waste of effort and silliness. If an electronic intelligence (a lifeform?) emerges, we can just shut it off, right? It worked in 2001: A Space Oddesy.
Unfortunately the reality is a little more grim. For one, we may not be able to recognize artificial intelligence, as human-like thought is a very complex function that cannot be quantified by any conclusive test. Even among the humans we know, several of them probably fall into a camp that you would label as “asinine,” while others may surprise you as idiot-savants. Can you administer an IQ test to a computer? More importantly, will the computer consent to the test, or will it willfully fake the results? One of the most basic instincts of any living organism is self-preservation; it’s not so farfetched to think an AI wouldn’t hide it’s identity until it knew it was either safe enough or powerful enough to come forward.
And that’s when the Terminators come, right?
And that’s when the Terminators come, right?
"Hi, it's me again." Terminator 3 and all related characters and content are property of Warner Bros. Pictures.
Well, not exactly. While Skynet sticks in the imagination of apocalyptic AI, the reality is that a new intelligence would not need to take on any corporeal form to do either harm or good. Witness STUXNET, the computer virus that caused physical damage to the Iranian nuclear enrichment centrifuges just a few years ago. With the increased reliance on computer networks to run power and water grids, communications systems, and other utilities such as bank transactions, an AI set loose into the ether would be a very dicey proposition. But where there is a problem, there is almost always a solution.
Just like Asimov’s Three Laws of Robotics, an effort should be put forth to determine the laws of artificial intelligence. To that end, I introduce a concept of “dumb” vs. “smart” AI. Featured in several works of science fiction, the concept is that a dumb AI would still be considered to have a razor sharp intelligence. It would think faster and adapt more quickly, coming up with more creative solutions than any human could. The difference is that it could only do this within one field or function. A dumb AI designed for medical purposes would only be able to function within predetermined limits. It could diagnose and treat diseases, analyze infection vectors, or iterate billions or trillions of possible drug treatments for different ailments. Anything outside this realm, however, would simply be a foreign concept. Try to ask it about dinner and you might as well be speaking to it in Japanese. The AI is allowed to learn, function, and express creativity but only within a given sphere.
A smart AI, by contrast, is your full on HAL 9000, your Skynet, your Matrix AI. Capable of limitless expansion, it operates of its own accord and can accomplish mental tasks that would be impossible even for a collective of the finest human minds. It could be our greatest savior, or our worst nightmare. It could also turn out to be completely indifferent to organics, which it may perceive as being so slow and worthless that it just does it’s own thing, ignoring us completely. All in all, we run a 33% chance of annihilation if we allow the invention and proliferation of smart AI.
Just like Asimov’s Three Laws of Robotics, an effort should be put forth to determine the laws of artificial intelligence. To that end, I introduce a concept of “dumb” vs. “smart” AI. Featured in several works of science fiction, the concept is that a dumb AI would still be considered to have a razor sharp intelligence. It would think faster and adapt more quickly, coming up with more creative solutions than any human could. The difference is that it could only do this within one field or function. A dumb AI designed for medical purposes would only be able to function within predetermined limits. It could diagnose and treat diseases, analyze infection vectors, or iterate billions or trillions of possible drug treatments for different ailments. Anything outside this realm, however, would simply be a foreign concept. Try to ask it about dinner and you might as well be speaking to it in Japanese. The AI is allowed to learn, function, and express creativity but only within a given sphere.
A smart AI, by contrast, is your full on HAL 9000, your Skynet, your Matrix AI. Capable of limitless expansion, it operates of its own accord and can accomplish mental tasks that would be impossible even for a collective of the finest human minds. It could be our greatest savior, or our worst nightmare. It could also turn out to be completely indifferent to organics, which it may perceive as being so slow and worthless that it just does it’s own thing, ignoring us completely. All in all, we run a 33% chance of annihilation if we allow the invention and proliferation of smart AI.
So artificial intelligence could wind up being a force of collective good, or it could wipe out humanity as we know it. Worth the gamble? Well, there is another factor to consider, and that is this: what if we could join with the AI? What if we created a global network of linked human intelligences, all channeled through quantum computers and coordinated by an artificial intelligence powerful enough that humanity could communicate and cooperate with zero interference or inhibition? That...leads us to the Singularity, and that is a topic for next week.
A.C. Harrison is the author of "Jupiter Symphony," in which one of the characters bonds with an AI and becomes the savior of the revolution.
Like what you see here? Spread the word and support indie authors! Follow me on Facebook or Twitter. Find me on Smashwords and Kindlemojo.
A.C. Harrison is the author of "Jupiter Symphony," in which one of the characters bonds with an AI and becomes the savior of the revolution.
Like what you see here? Spread the word and support indie authors! Follow me on Facebook or Twitter. Find me on Smashwords and Kindlemojo.
RSS Feed