The ultimate singularity is when a machine intelligence has emerged which can do anything humans can do. It might not do it in the same way, and it might need many robots to do its physical bidding, but it will be able to do anything humans can.
Moreover, the ultimate singularity will not have happened until this machine intelligence is capable of sustaining itself. In other words it will be able to re-build, reconstitute and re-boot itself in an infinite series of upgrades.
This then will mark a turning point. It will signal the transferring of the main driver of further progress from humans to machines. It will mean that humanity will now be second-best to a complex array of hardware, software and robotics.
This is the ultimate end-game for humanity, to produce a technology which is cleverer than it is, more intelligent than it is and which requires no further intervention from mankind at all, but which can renew itself though an infinite series of iterations.
What then will happen to mankind, once it has reached this epochal event? That of course is the great question, which cannot be answered, but only guessed at.
One possibility is what we might call the golden scenario. Our technology, now cleverer than us, continues to support the human race. It – of its own volition – continues to improve itself and take as its function the betterment and improvement of all mankind.
It might develop new drugs, design entire irrigation systems for parched areas, and grow new crop strains. It might forge ahead with research into new power sources, build better mass transit systems and streamline international trade. It might be able to employ its endless and all-powerful intelligence to marshal and collate the world's data, in all sorts of beneficial ways.
One part of this would be its ability to sift and employ all the information on the world wide web. At the moment computers can store the information on the world wide web, but they can't really understand it. Imagine if a computer could actually read and make sense of all that information – what an unstoppable force it would be.
We would certainly be dependent on this new intelligence to the highest degree. It would provide the software running our drinking water systems, our social media, our communications and transport structures, not to mention manufacturing and distribution. It would know literally everything about us, through our interaction with social media.
If cheap, lightweight robots became a reality, working in boring repetitive jobs, or doing the washing-up in people's homes, these too would be employ the same artificial intelligence running everything else. It will undoubtedly be the case at some time in the future that an array of software, hardware and robotics will come to underpin human civilisation.
Of course, the whole point about singularities is that you can't see beyond them. They are game-changing. They totally change all the presuppositions, realities and certainties which existed before.
There are many other scenarios, which are far less benign. One obvious scenario is that the artificial intelligence of the future simply loses interest in the human race, and switches off the water and the grid. This unpleasant prospect would set back human civilisation by at least a couple of thousand years, and lead to a far smaller number of inhabitants on Earth than at present.
Another possibility is that the future intelligence would decide to annihilate the human race entirely, though it is hard to see why it would do that. Why would it be necessary to destroy every last homo-sapiens on the planet, when simply switching off the power would send him back to the iron age, or perhaps even the Palaeolithic?
Behind all this of course lies a software and hardware supplier: a giant tech corporation of enormous size. I speculate in the book that there will be only one software provider – Avocado – which is furtive but benign. But the existence of commercial pressures, of competitive advantage and the need for commercial confidentiality all add to the uncertainties of the future. It might be that fail-to-safe systems, faults and flaws of tremendous significance are ignored in the race to be the dominant supplier of the software of the future.
Of course, we could try asking the super-technology of the future what its intentions were for the human race. But any machine intelligent enough to re-build itself, to re-constitute itself in an infinite series of iterations, without any further input from man, could effortlessly hide its true intentions.
Of course, it could originally be programmed to take the good of the human race as its number one goal; but how could we be certain it would continue to do so, if it was so clever? In practice it would be impossible to know what it was really thinking.
It is even possible that the technology of the future might decide it – in its then current form – constituted the pinnacle of the evolutionary process. It might decide to shut down not just the electricity grid and all life-support systems to humans, but also itself, as there would be no point in further development. In that case of course humankind would have to go back to the simple farming communities of 2,000 years ago – or even earlier.
So what is the future for mankind? It might be a bright and optimistic one, where work becomes optional and we can all head for the beach. Friendly but docile robots might bring us martinis, and engage us in polite conversation. Of course for those who want or need to work, no doubt jobs can be invented, for those with the desire to do something.
Fortunately we are still some way away from the ultimate singularity – perhaps two or three generations. The human race has not yet fulfilled its evolutionary purpose: to create a technology which is cleverer that it is, which is self-sustaining and which can improve itself infinitely, without any further human intervention.
What happens after this singularity is impossible to know. We can only hazard guesses. But of one thing we can be sure: that sooner or later this time will come, and that the human race will have to hand over the mantle of superiority over all else, to a vast array of software which is clever than it is.