Document created: 9th December 1996
Last Modified: Thursday, 21-Jun-2012 14:04:08 BST
Click here for Shadow home page
Back to Main Philosophy Page

Rich Walker on Why He Builds Robots...

I'm sorry, what did you say you were doing?

Building robots. Humanoid ones,


Well, like, houses are designed around humans, all the standard tools are designed for humans, humans are used to humans; I mean, you know the Daleks? Wheeled bases? How do you take over the Universe if you can't even climb a flight of stairs... Indeed -

No, I don't mean the humanoid bit, I mean the robot bit?

Well, someone's got to do it.

Yeah, but why?

Well, there's all these things I'd rather not have to do, so I thought I'd spend the best twenty years of my life slaving away over hot soldering irons, glue guns, keyboards and core dumps so we could stop doing them if we felt lazy...

You switched from I to We there?

Yeah. Do you like all the things you have to do everyday?

No, not really.

So, which don't you like?

Washing up, vacuuming, chopping vegetables, cleaning, - why, the list is nigh-on endless

Precisely my point.

But, wouldn't it be quicker and easier to produce machines that just did these things?

Maybe. But, two things. Would you use them, and clean them afterwards? Would you like to buy a new machine every time you thought of a new thing to do with them? It would be a lot handier if you had a machine that would make use of your existing tools to do the job for you, and would do it exactly as you wanted it done.

I see your point.

And, just think of the things we haven't mentioned?


Hold on, I've a list to hand...

But I haven't got any brass or silver!

Probably because of the length of time you'd have to spend cleaning it if you had it...

I see the point. So, how much will it do?

Weeeeeellll, have you heard of the Turing Test?

That thing about trying to tell a computer from a person by typing at them?

Precisely. Now, given what we've been talking about. Can you see anything useful about humans that the Turing Test doesn't test for?

Perhaps the ability to actually do something?

Such was exactly my thought on this matter. The Turing Test measures only the ability of the machine to process symbols; we require something more, well, concrete. Ideally, we would have a test that measures the ability of the machine to process not only symbols in a useful manner, but also objects.

Would I be right in supposing that you might just have such a test?

Is a bear - never mind. Yes, in fact, I do have such a test. I call it the Lasagne Test, and traditionally arrange for dramatic music at about this point.

Why dramatic music?

Oh, for that feeling of heightened anticipation, really; an overdeveloped sense of drama has always been amongst my chief failings. But anyway, the Lasagne Test goes something like this. Following a hard day spent doing whatever it is you do, you arrive home late, and tired. It's a cold evening, and the journey home was stressful beyond comprehension. You stumble in through the front door, to a picture of domestic bliss and tranquillity. The house is clean and tidy, a well-laid fire is burning nicely in the hearth, your slippers and a comfortable outer garment are placed handily by your favourite chair. A refreshing beer is just being poured by your robot; you sink into the chair, and begin to sip it. The robot passes you the paper, with the comment

I took the liberty of marking these articles I felt might be of interest to you; I also recorded this programme I felt might also be of interest. John, Dave and your mother telephoned; I told John and Davey! You might see them in the pub this evening; your mother and I discussed your Christmas plans; I did not commit you to anything.

You relax for a while, and decide to check the food situation. The robot pauses for a moment, and then says I hope madam will not be excessively displeased if I say that I took the liberty of preparing a lasagne earlier. To which, of course, you reply Splendid! Exactly what I wanted!

And how did the robot know what I wanted?

Simple; well, fairly simple. It knows what's in the kitchen. It knows your usual eating habits: the fact that you'll probably have eaten little or nothing since breakfast; the lightness of your breakfast and the stressfulness of the chill day, necessitating a large dose of hot food rich in energy. The robot has also a more-than-nodding acquaintance with your traditional eating likes and dislikes, and can quickly weigh up these factors to come to the logical decision: it should probably make a lasagne.

Good grief. Doesn't that require a rather large degree of intelligence?

No, I don't think so. It is a hypothesis of mine that things appear to be intelligent to the degree that they agree with you, and anticipate you . So, if we produce a machine that is very good at learning your patterns, and moulds itself around them, you will soon come to regard it in much the same light as you regard, say, a comfortable pair of slippers.

What, old and ready to be thrown out?


So, aren't you afraid that, with this ideal servant lying around, you'll cease to do anything useful with your life, and just sit around and be served hand and foot all day?

Not really; there's only a certain amount of time a human can sit around having orgies, going to gladiator fights, sleeping, drinking, watching TV. and so forth, before terminal boredom sets in.

And then what?

Frankly, given that history repeats itself, I expect to see an intellectual renaissance.


Look across history. Every period where there's been a new leisure class created has resulted in a great boom in the provision of things for them to do with their time. The very term Renaissance, of course, comes from the most famous of these periods. We can, I think, expect to see people engaging in all those activities we subsume under the mantle of culture, and doing them precisely because they feel it would be amusing. The world will be buried under a great sea of masterpieces! [foam gibber]

And what exactly do you mean by culture?

Hakim Bey has an elegant contrast that sets it forth rather neatly:

What we like about Palaeolithic life:

in short, culture.

What we dislike about civilisation can be deduced from the following progression:

  1. the "Agricultural Revolution"

  2. the emergence of caste

  3. the City and its cult of hieratic control

  4. slavery

  5. dogma

  6. imperialism

  7. [?]

I mean, Bey is generally regarded as a little out on the edge of the playing-field, but I think he hits the mark rather well...

Bit of an odd mark?

True. However, is it not a valid one?

Well, I don't know...

I think it is. After all, we are looking towards a transformation of our approach to what we do, and how we go about doing it. Our very approach to life will change!

How exactly?

Things that previously were large and important factors in our lives will suddenly become minor things drifting on the sidelines, taken care of for us on a day-to-day basis. True, we'll have the option of doing them if we want; if you feel like cooking an interesting meal, or rearranging your tape collection, you can do it; but the maintenance; ah, the maintenance can be left to the robot.

That's an interesting idea, using the robot to maintain a certain standard of living for you.

That's an interesting way of looking at it, too. In fact, the general functioning of the robot is to act as a homeostatic device on your reality. It attempts to maintain a significant level of negentropy, so you can go on worrying about the BigBigPicture without constantly remembering to deal with the trivia. I see this as quite a cheering prospect.

Yeah, the idea of having everything as you like it all the time is a pretty tempting one. Mind you, I've heard it said that these sorting-out activities are good for you, that it's important for your mental balance to spend time engaging in mundane trivia. What do you think on that?

I'll be interested to see if it's correct, basically. I think I could cope quite happily without doing these things. But this is the exciting aspect of the whole thing: we are not only engaging in an activity that could well make for significant changes in the way we conduct our lives, but also we are, for once, going in with our eyes wide open.

Didn't Edward Teller say something like that?

Well, yes, probably. On the other hand, science fiction has been discussing the nature of robotics and robots for years, and a lot of profound thinking has been done there, in a wide variety of fields. Many of the initial problems have been overcome: look at the Asimov Laws. The thinking that remains to be done is even more interesting than what has been done so far.

More interesting?

Yeah. How do you structure a society in which most of the menial labour is done by robots? Suddenly, the notion of 'full employment' becomes a little silly: the work that currently keeps most of our labour force in work is, clearly, entirely amenable to being classified as menial labour, and so disappears.

But... that's... silly... isn't it?

Oh no. Consider: no longer do we need to think in terms of 'we must find work for everyone'; instead, we get to think as 'we must find entertainment for everyone', which is far far more amusing.

Mass partying all the time?

Well, I suspect that all the time would be overkill. There will be things left to do, though.


Art. Science. The traditional pursuits of the 'gentleman of leisure' become available to all of us: we can simultaneously dabble in ten fields, with no-one to say 'where's a piece of productive work'.

Is there enough of that to go around?

It's true that, once you've eliminated the menial labour, the remaining work, the interesting work, is today being done by a surprisingly small segment of the population: under the current system, this work could not be extended to the whole population.

So, how do you propose to keep everyone busy?

Any attempt to retain full employment either has to follow the Luddite approach of saying 'but humans should do this, not robots, because it's good for them' or some such, or taking the more interesting and exciting approach of saying 'well, okay, we're going to change the way we do everything. Sure. Then let's think about it beforehand, and come up with an interesting method of changing it, to produce a system that we'll be happy with afterwards.'

You've managed to avoid the question.

Yep. I haven't explained how to keep everyone busy, because I don't know. If I could think of ways to keep everyone busy, I'd be Employment Minister. As it is, I'm open to suggestions.

And, anyway, what would you be happy with?

I've always thought that, given a plausible influx of money, perhaps around 10k a year, I could find I had no spare time, all the time.

But that's just you: that doesn't offer a way to deal with unemployment!

The problem with unemployment is twofold: firstly, people are trained from an early age to feel that they have to work in these constrained codified, censured environments, where there are externally-imposed schedules, and they do not have to think what to do: after all, you don't want that in most employment situations, because the employees who think are usually the ones who, at some level, cause problems by asking questions.

Didn't Plato want to ban artists from his ideal state because they were too prone to asking difficult questions?

'Fraid so. Of course, the other problem with unemployment is the fact that the money paid to unemployed people is designed to tide them over until they can get 'proper work'; if, instead, they were just given money and told to get on with whatever they wanted to, then there would be fewer problems.

Get on with whatever they wanted to?

Yeah. Bucky Fuller pointed out an interesting fact in correlation with this; from looking at leisure classes, he noted that if you gave ten thousand people money to live on, and told them to go away and get on with their lives, one of them would come up with something that covered the overall cost of the whole exercise. Ten of them would come up with something that covered about a tenth of the cost. He projected this down to the lowest level, which is, basically, the system pays for itself massively.

But what was he using as a measure of cost? How do you measure the benefits of someone's life?

I'd love to find a good measurement of 'standard of living'. I think that the measure that was being used was 'cost to the whole system': it requires x units of resources to support a person; but, by some action they take, they have improved these bits by this factor, and benefited people in these ways, which adds up to more than x.


Of course, the Daleks did in the end manage to develop a technique for climbing flights of stairs; indeed, when looked at from an objective distance, it was obviousness itself: they simply produced a mystical blue glow from their base, permitting them to levitate. We, sadly, do not yet have recourse to such tools, and so are forced to take more standard measures. Like legs.

I realise that not everyone likes lasagne; however, bear with me, as this is for illustrative purposes only. Feel entirely free to refer to this as, say, the Chilli Test, or the Roast Beef Test, or even the Nut Cutlet Test.

Within certain limits, of course; there are areas in which the robot should not intrude, and these would be known to it; there are areas wherein the robot will not be able to make any useful predictions, and these are for the most part the areas where, if it could make predictions, people would begin to feel threatened by the machine. The ability of the robot to be superlatively good at anticipating certain needs, whilst being wholly ignorant of others, allows us to produce what might be termed 'the perfect butler'. Discreet, accomplished, ever ready to help its owner.

Not, I hasten to add, that I necessarily advocate orgies; just that, looking at history, whenever there's a new leisure class, they do tend to spend a fair amount of time engaging in exactly that activity.

See previous note, substituting 'watching gladiator matches' for 'orgies'.

Alan Turing, noted British academic, mathematician, homosexual, cryptographer, and computing pioneer. He invented a test for 'machine intelligence' that tested exactly what Turing believed to be the significant part of intelligence; this reflected strongly, one feels, various period views of the supremacy of thinking over doing.

A homeostatic device is one that adjusts some things about itself in order to keep others the same. For example, a thermostat and a radiator make a homeostatic system: the thermostat adjusts the behaviour of the radiator to keep the temperature constant. In our example, the robot and the house form the homeostatic system.

Negentropy is the opposite of entropy: instead of things running down and falling apart if left alone, as would happen in a system with a significant degree of entropy, the negentropic system (house+robot) improves in quality if left alone. At least to a certain point.

Edward Teller: physicist, and 'father of the hydrogen bomb.' One of those things you really want your name associated with through the mists of the future: a particularly effective way of destroying otherwise-fruitful lands and peoples.

Asimov's Three Laws of Robotics:

  1. A robot may not injure a human being, or, through inaction, allow a human being to come to harm.

  2. A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.

  3. A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

Menial labour: that work which is currently done by human beings, but would be done perfectly well by robots if we had them; often used as a synonym for soul-destroying work, make-work, ?.

An image saying Millenium Products   Click here to goto Shadow Robot Company main website

©Copyright Shadow Project.

  An image saying Engineering Industries Association