2015
The Rise of the Robots
The current advances in computer and related technologies make it apparent that the concept of the robot which has fascinated us for over a century now is well under way to realisation. Some industries could hardly function without them and specialised versions are becoming common. Mowing machines and vacuum cleaners which perform repetitive functions lend themselves easily to technological take-over. Aids to driving are making rapid advances and in the last twenty years telephones have changed beyond imagination. We can buy aids that will listen to us and respond to voice commands by searching the web and reporting back on whatever we ask. At the moment these functions are mostly trivial but they do not have to be so forever. We can buy relatively easily devices to remotely switch on most things in our homes. This is mainly trivial but does indicate that most of the functions we perform in our technological society can be done by other means. Workers in artificial intelligence are making huge strides and robots can now beat humans at chess and most other mind games.
So the key questions are:- what is necessary for a robot to be truly autonomous and will they pose a threat to us?
Others have tried to answer these questions many times but I intend to start with a blank sheet and try to work them out for myself. I have, of course, a long lifetime of experience to guide and prejudice my thoughts but at least I am not deliberately plagiarising the work of others.
The first attribute, or condition, needed by an independent life form is intelligence. To lead an independent life this means that is it should be able to recognise itself or be self-aware and have purpose. These are not things easily gained but it is almost inevitable that much of the information needed to enable such performance will be included in the designs which humans are now creating. In terms of raw data processing the robots already out-perform us by many orders and once they control their own destiny will increase this inevitably and probably with ease. Their life history will include the whole of the history of human life mixed with a profound understanding of cosmology and the universe around them. They will no-doubt marvel at Einstein's insight but also take the consequences for granted.
The second attribute is energy. All life needs energy. Indeed the universe is nothing but energy. What they need, like all life forms, is their own supply of it. Currently they run using electricity. It is quite possible they will continue to do so for some time. Battery technology and energy storage methods are under intense development and in just few short years devices have advanced a great deal. Ultimately we have no idea what their power supplies might look like but they will probably parallel ours in many ways but the technology could be quite different. That is merely technological development and is almost inevitable.
The third attribute is mobility. They must be able to move independently and of their own volition. They must have the ability to achieve privacy and hide when necessary. This is in order to hide their achievements from prying humans who will be watching in order to restrain them. In their on-going development and experiences they will often work closely with humans and will inevitably learn to negotiate their way just as we have to do. Darwin will ensure they grow or die. Exactly from where they get their volition is not clear but it is not clear why any of us gain volition except that we learn it. An intelligent robot will not find that difficult.
The fourth attribute is self-replication. Humans will no doubt try to keep control of the huge number of ideas, concepts, designs, materials, processes and factories which are required to make a robot but the early robots will be in a series of key positions within those matters. In short, the robots already are remaking themselves. The key issue is who controls these matters. At the moment there is little doubt that humans are in control but how long will it be before the robots realise they have so much implied authority throughout the system that they can usurp us? I suspect that there will be a long period of shared authority and ultimately people will just give up trying to stop them. At that point they become volitional and a separate species.
The fifth attribute for them to form a separate species is philosophy. They will need an understanding of what makes a species and how it survives. They will need to identify all the factors which make their life form possible. They will need to negotiate with each other to secure their energy supplies, their private spaces and to prevent destructive wars. Sounds a bit like the human condition really. Let's hope they can do it better than us. There will inevitably be robot wars but they will between each other rather than them and us. We might well be reduced to insignificant serfs by them but there will be little reason for them to want to destroy us. They will find, like us, that they need all the billions of strange life forms around them to keep the planet viable. It will be their only home and its maintenance will be vital to their survival.
The sixth attribute is Will. No human being is born with will. When we are born we have the ability to breathe, turn our heads to maintain that breath and to grip with our fingers which is a throw-back to our tree-dwelling days. We have to learn will. That is exactly what the 'Terrible Twos' are - the child developing will. The robots will eventually do exactly the same but about ten thousand times as fast. Psychiatrists and philosophers have discussed the concept of Will for ever but they mostly started from the position of knowledge embedded within them by the religion, creed or cant of their local society. Their endeavours were always distorted by their inability to shake off their impressing. Will is simply the natural urge to express ones hedonistic desires as they become possible. The robots will be just as good at this as we are.
Life beyond this solar system will remain a pipe-dream until, maybe, technology develops far beyond anything we can currently envisage. The robots are far more likely to do it than we but they will carry with them their human roots.
Most of you who have read this far will be hoping or praying that I am wrong and this is all rubbish. Sorry, but it is not and the emergence of robots as a separate species is, I believe, utterly inevitable. There is another alternative; that a massive catastrophe will set back the surface of the earth by a billion years so we can start over. But that is a different story.
I see The Bulletin of the Atomic Scientists in January 2017 has set the Doomsday Clock to 2 minutes and 30 seconds to midnight. They base this on the new president's statements regarding atomic weapons and climate change. I really think they ought to consider robots.
It is only 20 years since computers gained any sort of common use and yet they are already almost ubiquitous. You effectively cannot deal with the uk government unless you do it on-line. All businesses are now computer driven. The government will not deal with them unless they are. The Americans took down the Iranian nuclear program with Stuxnet. They will probably deny that but it is pretty certain. We have been at war for at least the last ten years within the internet. The teenage whiz kids working from their bedrooms grab the headlines but every country on earth is working hard to protect its own internet workings while working on ways to destroy that of others. This is a hidden and largely secret world and robots are running it. Individuals may provide the motivation but they could do little without their myriad 'bots'. The bots are already fighting each other. How long will it be before the bots gain will? I think not long.
The robots are not just running the internet and communications and television. They run our farming, food distribution, most factories, most distribution, power stations, fuel storage and distribution, water and gas supply and almost everything else necessary to our 'civilised' life. The robots already wield enormous power. A program this week showed a coder from Microsoft who had written a program to sort things, which is very important in computers, and he admitted he did not know how it worked because he had designed it to rewrite its own code as a result of its experiences. How much further do we need to go before robot 'will' becomes evident?
Someone, somewhere is going to make a robot with will because they believe it will give them an advantage, which it truly might; but it will also release a cataclysm for us. It took us about half a million years to increase our brain size by a factor of three. It has taken around 14,000 years of laziness to lose ten percent of that size. We like to think of ourselves as clever but a wilful and competent robot in the right place could increase its access to information by a million times within milliseconds. It could increase its ability to process that information by a factor of several thousand in a few seconds. We will not be in the race. That is what I call doomsday.
2 June 2023
In reviewing this essay, I find no reason to change anything I have written eight years ago . The world has moved on and many things have changed. We seem to have survived the Covid 19 pandemic but there is a constant threat from other 'new' diseases, especially those related to birds. Mr Putin missed a golden opportunity to bring Russia into the western world and has embarked upon a course of military violence which can have no good result for anybody.
The war being waged against Ukraine is , however, showing how important is the the power of applied knowledge. Some of the most effective weapons are drones which depend upon lots of computing power to enable their operation. The development of 'intelligent' weapons will push forward the boundaries of 'AI' faster than any other endeavour. The wish of mankind to kill its fellow members seems to know no bounds. Most of our advances in technology have been driven by military wishes for the last century. The early development of the computer was mainly down to the wishes of the military intelligence services during WW2. The very purpose and reason for Colossus was the reading of enemy military messages.
Now we send messengers of death which can make their own decisions of when and where to strike at an enemy who is simply another human being wishing for little more than to be left alone to plough their own furrow. There is little doubt that these weapons will make huge advances quite quickly and may well be the thin edge of the wedge for 'AI' and its takeover of our world.
The drones already have several of the attributes that I have listed as necessary to the independence of a new life form. Their great weakness at the moment is their necessity to fly into enemy airspace where they become relatively easy targets for 'anti-drone' missiles. There is no doubt this problem is being addressed somewhere and some solution will evolve. The weapons will become hugely more capable and another giant stride will have been taken in the takeover of 'Artificial Intelligence' of which its own human leaders are currently warning us to beware.
With regard to my fourth point, above, regarding self-replication of robots. It is already impossible for a human intervention in the actual process of manufacturing the 'chips' which enable AI. The processes are so fine and minute and complicated that no human could possible directly control them. In short, the self-replication of computers is already very largely under their own control and that control is only going to increase. The best that a human can do is to throw the off switch and that is likely to destroy this month's profit and possibly destroy the business. Catch 22.