Would you let an evil Siri drive your car?
That's right. One day Siri might turn on us all. Or at least, it is a possibility.
The development of Artificial intelligence has got to be one of the most fascinating fields in all of science and technology today. To replicate human intelligence in a machine would almost certainly be one of the greatest accomplishments in human hstory, if not the greatest. It would open up a whole new world of efficiencies and abilities that we could never have imagined would be possible. The really interesting thing though, is that there are two vastly different futures imaginable with AI technology, one unimaginably good, and the other, unimaginable bad. Let me take you through them.
THE GOOD HYPOTHETICAL
The best part about the advancement of AI technology has got to be that we will all be doing less work. I mean, come on, who doesn't love the idea of just sitting back, relaxing, and letting your computer do everything for you. That would be pretty sweet. And as machines get better at more and more complex tasks, this reality becomes more and more likely. Think about it: we already have cars that can drive themselves! About a hundred years ago we did not even have cars, and now they can drive themselves, that's crazy!
Now, it is worth talking for a moment about exactly how fast these machines will be able to work. Consider a year on earth, the amount of time that passes, and what we are able to achieve. Obviously what we are able to achieve in a given year varies, but mathematically, there is an average of how much we actually complete/do every year. The neurobiologist and AI enthusiast Sam Harris projects that in the same fixed amount of time that a year presents, an advanced AI machine would be able to complete the equivalent of 22,000 years of human research and work. Think about how quickly we could solve problems like curing cancer and fighting malnutrition if we were actually doing thousands of years of work in a single year. Imagine how easy it would be to generate freelancing leads...
Now,
THE BAD HYPOTHETICAL
This is where things get juicy. The real jist of nailing AI technology is going to be replicating some sort of consciousness. This means creating a system that mirrors the human brain, so a machine would have our capacity for though combine with its computing power. This means a computer might be able to actually make decisions and think for itself as a human might. The hope is that this will create a renaissance of information, and launch us into a golden age of productivity. But the alternative is way less fun. There is potential for these machines to develop a belligerent and negative attitude towards people. This makes sense from a logical perspective, or at least, seems likely given what we know about the variability of human perspective. What would be the moral variability of the defining characteristics of a machines code of ethics? Will they like us? Or will they seek to end us. It is not a totally unlikely scenario that one of the first things an AI machine observes is our utter uselessness, or at least inferiority, in comparison to its mental fortitude.
A lot of famous people have theorized that this might be the case. The problem is, that WE DO NOT KNOW. It could be that there are fail safes that could be installed in AI software that would prevent a negative outlook on humanity. However, critics have raised the point that an AI machine would hypothetically, were it fully conscious, be able to override such imperatives in its programming. If this is the case, then there would be no stopping such a machine...
So what do you think? Are you terrified or gleeful at the prospect of an AI Future?
The development of Artificial intelligence has got to be one of the most fascinating fields in all of science and technology today. To replicate human intelligence in a machine would almost certainly be one of the greatest accomplishments in human hstory, if not the greatest. It would open up a whole new world of efficiencies and abilities that we could never have imagined would be possible. The really interesting thing though, is that there are two vastly different futures imaginable with AI technology, one unimaginably good, and the other, unimaginable bad. Let me take you through them.
THE GOOD HYPOTHETICAL
The best part about the advancement of AI technology has got to be that we will all be doing less work. I mean, come on, who doesn't love the idea of just sitting back, relaxing, and letting your computer do everything for you. That would be pretty sweet. And as machines get better at more and more complex tasks, this reality becomes more and more likely. Think about it: we already have cars that can drive themselves! About a hundred years ago we did not even have cars, and now they can drive themselves, that's crazy!
Now, it is worth talking for a moment about exactly how fast these machines will be able to work. Consider a year on earth, the amount of time that passes, and what we are able to achieve. Obviously what we are able to achieve in a given year varies, but mathematically, there is an average of how much we actually complete/do every year. The neurobiologist and AI enthusiast Sam Harris projects that in the same fixed amount of time that a year presents, an advanced AI machine would be able to complete the equivalent of 22,000 years of human research and work. Think about how quickly we could solve problems like curing cancer and fighting malnutrition if we were actually doing thousands of years of work in a single year. Imagine how easy it would be to generate freelancing leads...
Now,
THE BAD HYPOTHETICAL
This is where things get juicy. The real jist of nailing AI technology is going to be replicating some sort of consciousness. This means creating a system that mirrors the human brain, so a machine would have our capacity for though combine with its computing power. This means a computer might be able to actually make decisions and think for itself as a human might. The hope is that this will create a renaissance of information, and launch us into a golden age of productivity. But the alternative is way less fun. There is potential for these machines to develop a belligerent and negative attitude towards people. This makes sense from a logical perspective, or at least, seems likely given what we know about the variability of human perspective. What would be the moral variability of the defining characteristics of a machines code of ethics? Will they like us? Or will they seek to end us. It is not a totally unlikely scenario that one of the first things an AI machine observes is our utter uselessness, or at least inferiority, in comparison to its mental fortitude.
A lot of famous people have theorized that this might be the case. The problem is, that WE DO NOT KNOW. It could be that there are fail safes that could be installed in AI software that would prevent a negative outlook on humanity. However, critics have raised the point that an AI machine would hypothetically, were it fully conscious, be able to override such imperatives in its programming. If this is the case, then there would be no stopping such a machine...
So what do you think? Are you terrified or gleeful at the prospect of an AI Future?