The very moniker “artificial intelligence” alone should give us pause. Like “fake reasoning” or “fabricated knowledge” the actual term should conjure pictures of absurdity. Instead, sci-fi media and tech companies have made artificial intelligence something at which to marvel, aspire to and – if the movies are to be believed – feared. I think we’ve seen so many pictures of the “doomsday apocalypse” brought about by a bunch of robots realizing that the best solution for humanity would be extermination that it’s grown a little tiresome. I mean, really, how many times can Arnold Schwarzenegger swagger around helping puny humans against the onslaught of diabolical machines.
Artificial intelligence – or AI – however, is very, very real and in the last couple of years or so has made rather incredible strides. Unless you read tech blogs like I do, you’ve probably missed most of these, but I want to share them to give you some idea of what the rise of the machines looks like in our time. All of these things are truly amazing, some of them are a little creepy and many of them can be taken as harbingers of the future. Here are just a few highlights of things that have already been developed or are in the works. I’d love to hear your comments below.
Google’s Robotic Arms
Early this year, Google programmed a group of robotic arms to execute the mundane task of picking up blocks. They made it interesting, however, by connecting them together through a neural network. Over time, the robotic arms began learning better techniques to pick up blocks and taught the others using the neural network. They went beyond their basic instruction set, laid out by humans, improved upon it and then passed those advances on to each other. Yikes! If you really want to geek out, you can read the full report.
Late last year, engineers in the HRI Laboratory at Tufts University developed a robot that determined (on its own) to refuse to obey a human command. If the robot evaluated that the command put itself at risk (in this case falling off a table when commanded to walk forward). Despite repeated commands, the robot continued to refuse, until the human giving the command promised to catch it. While I can totally appreciate the need for the robot to not tumble to the floor, it also demonstrates a certain “awareness” that so far has not been seen. Stephen Hawking has some interesting things to say in regards to developments like this. It’s interesting that someone who believes that humanity is nothing more than an evolutionary hiccup should be concerned about our demise at the hands of our creations.
Just this week, news broke on Russia’s “Ivan the Terminator” remote controlled battle robot. While I almost always take news about Russian technological developments with a grain of salt, if even half of what they say is true, it could launch an “AI race” similar to the nuclear arms race from decades ago. This “terminator” is nothing more than a drone robot in the field controlled by a human far from battle. However, the way in which they’ve designed the interface – to mimic human movements in such detail that it can drive a vehicle, for example – opens the door to being able to program the robot with much greater “natural” movement, especially given the learning demonstrated by Google’s robotic arms.
Printing Body Parts
3D printing has been around for a while now. We’ve managed to print soap dishes, hammers, jet engine parts, entire buildings and even food. Now, we’ve taken 3D printing to a whole new level: living tissue. The researchers at Wake Forest University have literally mixed living cells with a gel that hardens to the consistency of organic tissue. What’s even more, this gel contains tiny “microchannels” where blood vessels begin to grow within 24 hours to deliver nutrients to the cells deep within the constructed tissue. While this doesn’t exactly fall under the artificial intelligence category, I can’t help but have images of Arnold’s Terminator robot clad in actual living tissue as I read this article.
Pain can be a great thing, right? I mean, without pain, we’d never know to take our hand off a hot stove or react to the stupid Lego pieces left haphazardly on our floors in the middle of the night. Robots don’t feel pain, which is why we can send them into some of the most dangerous places on the planet without thinking twice. But what if you wanted a robot to preserve itself like we do? You teach it to feel pain. Researchers at Leibniz University of Hannover in Germany believe that by programming robots to feel pain and react to it like humans do, they can effectively preserve themselves from danger. Since one of the main functions of robots in many applications is to do the jobs that are too dangerous or hazardous for humans, I’m a little unclear on the goal here…
Assembling the Pieces
So, let’s just stop and think for a moment about putting all of the above things together:
- Robots operating with a neural network both internally and between each other.
- The ability to feel pain and make decisions to preserve themselves.
- The ability to evaluate their circumstances and environment and refuse a human command.
- The ability to execute fine motor commands.
- 3D printed living tissue
Looks like we’ve got the makings of a flesh clad, autonomous learning machine. All of these technologies are available now and are being developed at an astonishing rate. What will this look like in ten years or twenty? We are at the point where all those science fiction movies have the potential to become reality. I’m not saying the robot apocalypse is imminent by any means (especially if Siri, Cortana and Google Now are any indication), but this technology is poised to radically change the world around us. Are we ready for it? Only time will tell.