Letter #9: It’s alive!…well, not yet.
January 21, 2016
Dear reader,
If you don’t already know, I recently contributed to Medium’s #PostsFromTheNearFuture campaign, where the top publications on Medium write from an imagined moment in 2016, looking at something they are excited by, concerned by, or interested in for the year ahead. My contribution explored the future possibilities of Affective Computing in our society. You should read it.
To write this piece, I had to look at what has already been accomplished by experts in corresponding fields of modern computation — quantum computing, neural networks, machine learning — and it was fascinating. To think that these machines have come this far because of our own intellectual achievements is a major feat in itself.
In this installment, discover everything you need to know about the current state of computer evolution and what it could mean for the future.
[Video] Something that learns.
The ability to learn is the building block to understanding, which drives all of our progress and evolution as human beings. Right now, machines are going through something similar — an evolution of intelligence. Machine learning, aka artificial intelligence, as explained by Google in this video, is an effort to build machines that learn from their environment, mistakes, and from people. More literally, it explores the study and construction of algorithms that can learn from and make predictions on data.
Machines that can think does sound a little concerning, especially to those expecting a skynet-esque outcome, but I think we’re pretty far from that at this point. Besides, we use little bits and pieces of AI everyday, to navigate cities, search the web, recognize speech, organize photos, interpret language, etc. This machine learning effort is taking everything computers are currently capable of, and putting it together to create something more unified. We’re reaching the stage in our own intellectual evolution where we’re going to need smarter computers to help study, analyze, and conclude solutions to major issues and dilemmas created by, or just fundamentally not understood by, our own species. We need computers that can think for themselves and be the unbiased middle man, or rather, machine.
[Article] Something smarter (and possibly on LSD).
In order for us, as humans, to learn, we need a brain to absorb presented information and turn that information into conclusions based on what we were presented. This is no different for machines. Google, and many others, are in the process of developing the digital brain by way of neural networks. Neural networks are what make machine learning possible, though not without human guidance. We write the algorithms that cause these machines to “think” and teach them what to look for―like a parent teaching their child right from wrong. These networks work in layers, communicating amongst each other to reach an ultimate outcome. Google is currently testing the limitations of these networks with methods such as inceptionism and iteration, through an effort called Deep Dream, and producing some really weird images while they’re at it. These techniques “go deeper” and peer into the “subconscious” of the computer to help determine how neural nets work to classify and remember information as well as organize its own neural architecture as it processes. This effort plays an important role in determining the potential of neural nets in future technology. You can also, most certainly, consider neural networks and Deep Dream valiant efforts to understand the way our own human brains work. Something humans have obsessed over since we realized we had a brain. What better way to learn than to recreate?
Additional:
- [Article] Google’s official blog post explaining Neural Networks.
- [Video] Explore Machine Learning & Deep Neural Networks with two Google employees.
[Video] Something more powerful.
With every technological advancement, it’s a given that the hardware becomes smaller and more powerful. So small in fact, that today’s computer components are nearing the size of an atom. If physics is preventing us from going any smaller, how can we continue advancing? The answer: quantum computing―a form of computing that researchers have been theorizing about since the 1980s, until companies like Google, D-wave, and Microsoft, started making it a reality. In 2015, some deserving light was shed on quantum computing and the possibilities of next generation of computers.
What is quantum computing exactly? Let’s just say it’s an au-naturel way of computing. By harnessing the natural laws of mathematics and physics we have an opportunity to create much faster, powerful, accurate machines. Ah, the wonders of science. But, because of this natural approach, we might have to throw out everything we know about classical computing―coding, hardware, operating systems, architecture―and start over from scratch. There’s no guarantee that what we know now will work with what we are able to develop. We can only speculate that quantum computing means many good things for our future. Though we have no idea what those good things may be, we still have a duty to explore and discover the possibilities of our universe.
Additional:
- [Video] Google’s official explanation of quantum computing.
- [Audio] Dive deeper into what to expect from quantum computing.
Even with all the knowledge we have now, we have no idea what the future has in store for us. For all we know all this machine learning hubbub might lead us right into a dystopian, AI dominated, skynet-type future. Or not. Maybe it’ll lead us to a perfect future, or more reasonable outcomes like major breakthroughs in the understanding of ourselves, reality, physics, time travel, etc. Who knows? But, I believe that we shouldn’t let our vast imaginations and fear of change get in the way of our progress (though it doesn’t hurt to be prepared for the worst). Otherwise we’ll never know what’s possible or what could have been. Regardless, the future looks promising as we look to understand the unknowns of our universe with a little help from some hyper-intelligent robotic companions.
Best,
The Internet Traveler
This is an adaptation of Letters from an Internet Traveler, a newsletter comprised of thought-provoking tidbits discovered on my travels across the Internet, casually delivered at least twice a month (maybe).
Subscribe here and follow the Medium publication here.
