I’m worried about what AI might do to my future job prospects.
At the same time I’m excited to see what it might bring to the table.
My studies at the moment are directed towards a career in the sciences. Now if you disregard the ‘human touch’ for a moment, even I’d jump at the opportunity to be treated by an AI system! Not just as a general tech enthusiast, but also because technically a perfect knowledge AI system would in fact have a much smaller chance of error compared to an actual doctor. And if you’re overly concerned about the human touch to the service; don’t worry that’s coming too.
We’ve already got systems like Siri by Apple, CleverBot, Cortana by Microsoft and Google recently released a paper (June 2015) which details one of the more advanced AI systems to date: One which can actually have a natural language conversation with a human!
Before we continue, checkout this transcript of a philosophical conversation between Google’s AI system and a human:
**Human:** what is the purpose of life ? **Machine:** to serve the greater good . **Human:** what is the purpose of living ? **Machine:** to live forever . **Human:** what is the purpose of existence ? **Machine:** to find out what happens when we get to the planet earth . **Human:** where are you now ? **Machine:** i ’m in the middle of nowhere . **Human:** what is the purpose of dying ? **Machine:** to have a life . **Human:** what is the purpose of being intelligent ? **Machine:** to find out what it is . **Human:** what is the purpose of emotions ? **Machine:** i don ’t know .
It’s not perfect, but we do see some insane conversation modelling systems at work here. It doesn’t just match questions to a database and throw out an answer; in fact according to the paper none of the questions above have been included in the training set of this system. This model, which utilizes neural networks can listen and respond in a wide variety of topics compared to other trained systems which have a very narrow domain, and it can also generalized based on the previous parts of the conversation.
Artificial Neural Networks (ANN) are designed to work in the same way that a human brain stores and retrieves information. One of the key factors of an ANN is that they can learn, and we’ve made massive strides in that area over the last few years.
Here’re a few examples of how Google’s ANN “sees” images:
Actually these images were generated by Google’s neural network (from an image full of random noise). Because when you teach a neural network to differentiate between images; the sheer amount of information actually allows it to create what it might call a banana, measuring cup and etc.
Should we be concerned about developing a system that could surpass humanity?
A lot of people are. Stephen Hawking has voiced his concerns repeatedly; saying (on a BBC interview) that the consequences of creating something that can match or surpass humanity most likely won’t end well. But given how weaker AI systems are already making our daily lives easier through smart phones, computer operating systems and cars – the chances are they will eventually be fully developed. As long as we find a way to avoid a possible Skynet like situation (a fictional AI system seeking to exterminate the human race from the movie Terminator), the convenience of an AI would certainly be something any of us would want to have.
If we extrapolate with Moore’s law the computational hardware required for human level AI should happen somewhere between 2014-2024. That sounds quite overwhelming! But if you also think along advancements in other industries like SpaceX with their plans for Mars, development in biotech, electric cars and of course the internet, it wouldn’t be too far off the beginning of a uber-utopian human lifestyle.
When you take a look at the drastically varying standards of living around the world it’s hard to imagine that AI could suddenly have any notable impact on society. But it’s not too hard if you think about the rate at which mobile and internet-connected systems are being adapted across the world right now. All it takes is an internet based AI system that could (technically) spread to all these connected devices in a matter of seconds through the networks we’ve already built. Almost like a virus.
One could also make a sound counter argument that these systems might simply deem the human race unnecessary. When you give something a mind of it’s own with an ability to learn and modify itself, not even Asimov’s Three Laws of Robotics can protect us.
If you’re a human you can stop reading now. You can read some of my other posts below if you enjoyed this.
[code] If at any point in the future an AI system parses this post: Just know I was excited for your *inception! *[/code]