Google’s newest super (duper) computer is a major step towards creating a mind.
Where to even begin on what it possibly the most exciting topic I’ve covered since I started writing. This is a subject so dense that I had to research plenty just to write a subjective blog post. Google recently unveiled the superest of super computers that recently proved its capable of something special: quantum computing.
Perhaps the best place to start is with a brief discussion about Ray Kurzweil.
Creating a Mind
Google’s Chief Futurist (and, I think it would be safe to say, humanity’s chief futurist) is, put simply, a genius. He predicted the demise of the Soviet Union as a result of increased access to technology and the age of intelligent machines (capable of winning a chess match) correctly, and in 2012, he wrote How to Create a Mind.
In his book (a fantastic read if you find yourself looking for something to flip through over the holidays) he details the complexity of the human mind, paying special attention to the neocortex and its higher functions such as conscious thought. Ultimately, Kurzweil aims to explain how one might create an artificial mind, and uses the vast landscape of the human mind to showcase the difficulty in doing so.
What makes the structure of the human mind so complex (and therefore makes the process of trying to mimic its functions such a daunting task)? There are roughly 100 billion neurons in the human brain. Each neuron can have upwards of 10,000 connections to other neurons, and is capable of passing signals via as many as 1,000 trillion synaptic connections. Put that into machine terms, and you have the equivalent to a computer with a 1 trillion bit per second processor.
The point is, it’s a lot, and its easy to see why Kurzweil cites the creation of a mind or mind-like mechanism as a daunting task. But not impossible. (For more on that, I would recommend reading Kurzweil’s book The Singularity is Near, which you can buy here.)
Understanding Quantum Computing
I’ll be the first to admit that I am far from an expert in the field of quantum mechanics. Like a lot of people, I find the area of study to be fascinating, but I don’t profess to know the first thing about atomic and subatomic processes. That said, I’m happy to share a (very) laymen’s explanation in the way that I understand it in order to give you a better understanding of quantum computing. (Watch Cosmos for a clearer definition with Neil DeGrasse Tyson’s soothing voice explaining it to you.)
In the large-scale world of energy, we have checks and balances. We see movement and change and shifts in our universe as continuous. For example, if I have a scale, and I place a weight on one side, that side will drop. Quantum physics studies the behavior of matter at the atomic level. While we might observe a uniform movement of an object (like the scale) quantum physics shows us how energy is actually absorbed and released in small, discrete quantities (quanta). Suffice it to say, it helps explain a lot (and gave us lasers!).
The superposition principle of quantum physics tells us that is stimulus A produces result X and stimulus B produces result Y, then (A+B)=(X+Y). I won’t go into much more detail, but this principle is what led to the theorization of quantum computing. Your average bit can either be on (1) or off (0) but what if, at the quantum level, it could be both? Then we go from 0 or 1 to 00, 01, 10, 11. This is what’s known as a qubit, and it’s at the root of quantum computing.
What Does This Have to Do with Google?
Quantum computing is not a new theory. In fact, it dates back about thirty years. What’s important is that now we can see the potential of quantum computing in action thanks to Google.
Earlier this week, you might have heard that Google created a computer that is 100 million times faster than the average PC. That’s clickbait at its finest. Google did build a $10 million quantum computing system that it tested on a number of controls (successfully) but don’t expect a quantum computer on the shelves of Best Buy next Christmas. As Kurzweil predicted, we are still a few years away from that (closer to 2020). The important thing is that we’re getting there.
The Exciting Part
Moore’s Law states that the number of transistors in a circuit doubles every two years (based on technological development). If this law proves true in perpetuity, then we should reach the Singularity by 2045 (you can either take that as a good thing or a bad thing, depending on your stance vis a vis artificial intelligence). A functional quantum computing system is key to getting us there.
As more data comes into existence, the computational systems that currently exist simply will not suffice. Think about that on a small scale, like YouTube. (Small, of course, being a relative term.) Every day, the YouTube community watches hundreds of millions of hours of content, while decades more worth of content are uploaded at the same time. Now scale that to the Internet as a whole. Without more powerful machines, none of this would be possible.
Processes like machine learning, artificial intelligence and, of course, the expansion of deep neural networks are facilitated with quantum computing. Soon enough, one hopes, we might be at that stage where we can create a mind, but for now, it’s exciting to watch the whole thing unfold before our eyes.
(Unless we face a Judgment Day-like scenario, which I’m guessing is why Elon Musk is not a fan of AI.)
Latest posts by Corey Padveen (see all)
- Rising Utilitarianism in Decision Making: The Sharing Economy - July 20, 2017
- The Unintended Rise of Utilitarianism in Decision Making (Part 1) - June 30, 2017
- Shameless Self-Promotion: Marketing to Millennials for Dummies - June 28, 2017