All Hail, our Robot Overlords!

Summer Glau as the Terminator. Fiction or reality?

You KNOW it’s going to happen! We can either extend our own biological capabilities by tapping into the science and innovation fuelling robotics and AI, or we’ll essentially become irrelevant.

I’m not the only one saying it. Ray Kurzweil’s been on this bandwagon for a while, in fact, I think he’s behind the wheel! Hans Moravec, no mental slowpoke, is there too.

Robots being developed today are capable of interacting with a world that was designed by humans, for humans. Stairs make sense for a biped, so robots are taking the form of stair-climbing bipeds. They can now run, with both feet off the ground at the same time. They can self-stabilize when you push them. We are not talking about Robby the Robot. What technology has achieved is far more impressive.

As the video below shows, Honda, is investing tons of money replicating the physical dexterity and implementing the basic instinct that has only been found thus-far in erect bipeds (e.g. you and me).

Computer vision is improving day-by-day. Improvements in ‘planning’ algorithms, combined with vision and fuzzy controls have yielded cars that drive themselves pretty effectively, at highway speeds:

And we also have advancements in brain scans, brain-silicon interfaces and mind control of electronic devices. That last piece has become so ludicrously easy to implement in its simplest form that the last Toy Convention held in Mahattan actually featured a mind-controlled toy! Here’s a video of that:

We already have voice recognition approaching 99.9% accuracy. Not theoretical, this. I’ve tried it. Pretty darn good. And of course, we have conversational systems that come fairly close to passing the Turing test, whereby during a 5 minute conversation, at least 30% of human participants would not be able to distinguish the computer’s conversation from that of a human. And we’ve made robots with skin actuators and, frankly, that look pretty… well… pretty:

We also have Stephen Wolfram and his Wolfram Alpha, a natural language driven system that uses the Web as it’s brain, ‘understands’ the knowledge that web pages contain and answers questions you might ask it.

Now imagine Asimo, in the body of that Japanese female android… with the car driving program embedded in it’s head, so it can drive… with Wolfram Alpha and an “almost passed the Turing test” conversational system as its interface to a human… with gesture, vision and speech recognition operating at 99.9% levels. Now, add to this the ability to store terabytes upon terabytes of visual, auditory and tactile data gathered via sensors as this being moves around. Add to that a Google-style search engine driven by – once again – speech recognition. Extremely powerful exo-skeletal technology that provides near-human levels of dexterity mated with superhuman strength. Since we are imagining, might as well throw in arts/music capabilities, theorem proving capabilities and some chess skills, for good measure. Sprinkle in some extra-sensory perception (at least as per the human definition), such as radiation sensors, pressure sensors, ultra violet and infrared cameras, wind speed and direction sensors, magnetic field sensors, ultra-sonic hearing, microwave and RF hearing/sensing, “Radar”, “Sonar” and laser-based vibration analysis to recreate sound, essentially, a “spy ear”.

All of this is available today. It’s just been implemented in pieces, independently. Yeah, the integration would be a bit of a challenge, but not a greater challenge than coming up with the component technologies were to begin with.

If such a “robot” existed, how long would it be before we stopped thinking of it as a “machine”? Would it be more interesting to interact with, than, say a dog? Don’t we anthropomorphize robotic vaccum cleaners already? Would these amazingly powerful creations be any different? Of course not!

I don’t want to make this post an AI discussion about learning, neural nets, support vector machines, fuzzy response systems and all that. I think the point is illustrated well enough with all these pictures and videos… and a little bit of imagination.

If we are willing to use Predators and Reapers, or a variety of UGVs as near-autonomous “soldiers”, what about our imagined Robot? How autonomous will we make these creatures? How smart will they get on their own, by being mobile and observing what we observe – only much better, at higher fidelity and with unlimited, instant-recall memory. What we’ve imagined, is this the Terminator? Science fiction from 20 years ago is becoming science fact, pretty fast. For those of you who are worried about the coming reality, here’s something by Bill Joy, the founder of Sun, that should scare you even more. Suffice to say, no one is going to heed Bill’s advice. Including me. It’s just not the way things work around here.

All Hail, our Robot Overlords!