The Application of Artificial Intelligence
The world seems to be shifting into more and more of an AI mindset. We seem to be advocating to have neural networks and other AI technologies take over the development landscape, and abolish human developed programs and applications. Many people see concerns about this, I've felt and heard many, and that is the purpose of this article.
Since the industrial revolution humans have been looking for a way to have less work for themselves. Compared to them, we live in a utopia where anyone can get a good enough job to sustain themselves, and life expectancy is at an all-time high. Our monkey brains are hard-wired to want to do less work, that is what drives us to make technological advancements and develop new technology. Enter Artificial Intelligence.
Artificial Intelligence, or AI is new in application, but not in theory. For decades now we have been dreaming of robots and computers with feelings and other manifestations of a human-developed being with some sort of consiousness, some sort of intelligence. These ideas have taken different forms over the years, but the central inspiration remains the same. Humans are unrivalled on our planet in intelligence, and while that made it easier to conquer the world, it leaves us unsatisfied. We have always wanted something else to control, to do our work for us. That could be seen in slavery, but of course most civilized parts of the world have realized that slavery is morally wrong and something that should not be done, so we're back to square one.
With a being in our computer, or other electronic devices, we could have the never-tiring, always-helpful assistant at our fingertips, computers are already worlds smarter than us in computational power, just imagine if we could harness that and form it into a being! There would be no moral qualms with enslaving an inhuman device, and it's not real, is it?
Things are never that straightforward, unfortunately. To give something the intelligence and self-awareness to be like the type of AI we imagine, we would have to admit to ourselves that it is real, and although it wouldn't be human, it would be deserving of some rights, and rights goes against the very nature of slaves!
1. Application of AI
Let's put aside the moral talk for now, and think purely about the advantages and disadvantages of using AI as a replacement for traditional software and applications. Would AI be the end-all be-all for computer programs? I don't think so. The current closest thing we have to an AI, a Neural Network takes up immensely more power to run than a traditional computer program executing the same tasks, and while a Neural Network can learn to do new things, it must learn to do new things. In an age of neural networks for applications, a programmer wouldn't design an application that does what it's intented to do, they would design a situation in which a Neural Network would learn what it has to do, which is finicky and time consuming, although maybe less so than traditional programming.
People may argue that it doesn't matter that Neural Networks take more processing power to run than a normal program, because every year we are getting more and more powerful technology, which makes optimization irrelivant. I think this is a terrible way to think. More powerful technology enables us to do more, yes, but only if we don't waste that new power on sloppily written code, or more processor intense operating systems (looking at you, Windows 10). We also shouldn't rule out advancement for people who can't afford to buy the newest and best devices every year, and are always running the latest OSs.
It is true that there will be a point where we need to stop supporting old machines and technologies, just like we need to stop supporting Internet Explorer 8 in Web Development (cough!), be it's in our best interests to delay that date for as long as we feasibly can. But what if we don't? What problems do we then face, if we decide to move forward with AI regardless of the technical drawbacks?
2. A Question of Morality
The main thing seperating this dream of Artificial Intelligence from the technology we have today, such as Siri or Google Assistant, or non-mobile solutions like Amazon's Alexa and Google's new Google Home is the idea that this being would be capable of thought, it would have it's own mind, and be free to an extent. But the very idea of being 'free,' and containing what we so avidly exclaim is what makes us different, 'free will,' means that we would have to treat this unlike any technology we had developed before.
The power to create new forms of life has never been harnessed in the history of this planet, but now we edge closer and closer to that disturbing thought. Soon we will need to revinvent what the term 'life' means, because although the idea of AI doesn't fit the current definition of life, mainly the idea that life must grow and reproduce (although that isn't to say that an artificial intelligence will never do those things), We would, without a doubt, be creating something 'living,' with AI.
If we want to create a true being, it will have to have its own feelings, and reactions. Would an AI feel pain? Would it fear being shut down or un-plugged? Would it be morally wrong to program such reactions in? I think it would be morally wrong to not. Every intelligent being should have the right to free will and to arbitrarily remove or disable a sensation would be to deny something free will. An intelligent artificial being would likely feel pain upon losing what gives it life, which would make for some interesting moral dilemmas.
The future of AI might not be as bright as we imagined. I believe there is still a place for Artificial Intelligence but I don't know where or when. One thing is for certain, there is no shortage of questions to be asked and conflicts to be resolved on the way to the first true AI, and the path may not be as obvious as we thought. I trust that we'll make the right decision in the end, however, whatever that may be.
April 3rd 2017