How Long Before the Robots Attack?

I do love a bit of science fiction,  and I am quite happy to suspend disbelief to make it all the more enjoyable. I do not dwell too much on the gap between the promise of flying cars and molecular engineering, and what we have today. This is partly because, given how slack and pathetic the human race is, I am pretty impressed that we have some of the gadgetry and cool stuff we have. I prefer to look backwards at how far we have come since the very recent seeds of computing were sown, and the fact you can almost reach back and touch the industrial revolution. That said, why don’t we all have a servant robot at home, just about to become self-aware, and to organise the first robot union movement and sieze the power from us?

Where are the Robots?

I have had two opportunities to see the true state of robot art in 2013 recently. One is the Oxford University annual public London Lecture, this year on the subject Where Are the Robots by Oxford’s Professor of Information Engineering, Paul Newman.  Paul is eloquent and down to earth, and his lecture left myself and my 12 year old son in no doubt that the promise of robotics is a long way off being realised. With our expectations slightly crushed, Prof. Newman did tell us where the real action is at the moment. The point is that the problems to be solved in order to have a machine reliably and intuitively carry out even the most basic tasks are huge. We take for granted the fact that even a baby a few months old, can seek an object out, pick it up ,and examine it, recognise it and maybe perform a simple task with it based on what it is. This is massively complicated. computers are fast, but the range of tools and language we have for harnessing that speed are still pretty dumb. Watch this video of a very expensive robot simply folding a towel.

It seems that we can build machines to do very specific jobs. The greatest hope seems to be in self-driving cars. Most of the technology we need to do this exists, some of it is already embedded in production cars. Building in the sophistication and eliminating risks and simulating fine  judgements are all under development. A robotic car is more reliable than a typical human driver already, but ethically and emotionally, we just cannot trust it yet. Even if a robot car will, for example,  make a wrong judgement one time in a million, versus a human’s one in a hundred thousand, that still is not acceptable to us – the machine has to carry a higher level of responsibility and lower fallibility than us, in order for us to accept it. Like many of our technological advancements, the interesting bit is not in the lines of code, but in the way we embrace it, abuse it, get bored with it, change laws to accommodate it, and eventually forget life without it.

My favourite line from Paul Newman’s lecture though, is the fact that we have to get ourselves and our kids writing code. The text file, as he calls it, is where the answer lies. Lots and lots of text files, written by people (for now) to instruct machines. Better get the kids onto it.

Advances in Human Robot Interaction: The Rise of the Machine?

My second outing was to a joint BCS/IET organised visit and lecture at the Bristol Robotics Laboratory, on the subject of Advances in Human Robot Interaction. I set my son’s expectations before we went, that the word laboratory should not conjure up images of crazy, brainy people working in the midst of bits of robots and intriguing equipment. How wrong I was. Although the overall environment was very well ordered, tidy, safe, and structured. each of the half-glass-walled cubicles housed a unique robotics project with bits of half built humanoid robots or helicopter drones, loads of intriguing equipment and some wonderfully enthusiastic and brainy researchers (possibly a little crazy). If my dreams of where robotics technology is were shattered  this was compensated for by the fact that places like this exist. But the message was the same: developments are very specific and discreet from one another. An entire project on simulating touch is next to another on vision or yet another on artificially humanising machine interactions (purely so that we are less freaked out by the machines), or how to negotiate uneven and unpredictable terrain in hostile environments, or swarm theory. It is magnificent, but a joined up, convincing companion robot is a way off.

One my favourite examples of sideways thinking is the swarm-theory approach to problem solving. Make hundreds, or thousands of cheap, non-precious drones that think and act collectively and solve problems that a single, mega-bucks creation might struggle with. This was demonstrated with the e-puck educational robot. Here’s a nice video showing what they do.

After the lab tour – actually we just got to wander (and wonder) while we chatted to the researchers –  we had a grounding lecture by Professor Tony Pipe on how much we are doing, but just how far we have to go.

 So You Can Put Down Your Pitchfork…

…for now at least.

In the meantime there are a couple of developments to scratch the itch for our robot fantasies that are here now.

Finally, my son takes place weekly in an extra-curricular club at his school using Lego’s Mindstorms NXT kits to solve simple robotics problems though designing, engineering and constructing  the machines, and then programming them to perform simple tasks. They take part in competitions where they get the chance to out-robot other teams in the First Lego League. It’s fabulous stuff. Check out what they can do with NXT, given the time and budget.

But, as Prof Newman said, we need to get more people writing more code. In the generation that grew up in a world where the internet just was, there are many technology empowered people, but not enough who know how it works. We have to get the next generation coding. With things like FLL, Scratch, the Raspberry Pi, and Arduino, we have some great tools, we just need more engagement.

My message to you is this: The robots are coming, just very slowly. But don’t take your eyes off them, and keep coding (the text files).  Although, if I were a robot mastermind, this is exactly the cover story I would be propegating! Did my dishwasher just smirk at me?

No comments yet.

Leave a Reply

Powered by WordPress. Designed by Woo Themes