AI is mainly a stimulative technology that will create jobs

AI has been getting a lot of bad press the last few months from doom-mongers predicting mass unemployment. Together with robotics, AI will certainly help automate a lot of jobs, but it will also create many more and will greatly increase quality of life for most people. By massively increasing the total effort available to add value to basic resources, it will increase the size of the economy and if that is reasonably well managed by governments, that will be for all our benefit. Those people who do lose their jobs and can’t find or create a new one could easily be supported by a basic income financed by economic growth. In short, unless government screws up, AI will bring huge benefits, far exceeding the problems it will bring.

Over the last 20 years, I’ve often written about the care economy, where the more advanced technology becomes, the more it allows to concentrate on those skills we consider fundamentally human – caring, interpersonal skills, direct human contact services, leadership, teaching, sport, the arts, the sorts of roles that need emphatic and emotional skills, or human experience. AI and robots can automate intellectual and physical tasks, but they won’t be human, and some tasks require the worker to be human. Also, in most careers, it is obvious that people focus less and less on those automatable tasks as they progress into the most senior roles. Many board members in big companies know little about the industry they work in compared to most of their lower paid workers, but they can do that job because being a board member is often more about relationships than intellect.

AI will nevertheless automate many tasks for many workers, and that will free up much of their time, increasing their productivity, which means we need fewer workers to do those jobs. On the other hand, Google searches that take a few seconds once took half a day of research in a library. We all do more with our time now thanks to such simple AI, and although all those half-days saved would add up to a considerable amount of saved work, and many full-time job equivalents, we don’t see massive unemployment. We’re all just doing better work. So we can’t necessarily conclude that increasing productivity will automatically mean redundancy. It might just mean that we will do even more, even better, like it has so far. Or at least, the volume of redundancy might be considerably less. New automated companies might never employ people in those roles and that will be straight competition between companies that are heavily automated and others that aren’t. Sometimes, but certainly not always, that will mean traditional companies will go out of business.

So although we can be sure that AI and robots will bring some redundancy in some sectors, I think the volume is often overestimated and often it will simply mean rapidly increasing productivity, and more prosperity.

But what about AI’s stimulative role? Jobs created by automation and AI. I believe this is what is being greatly overlooked by doom-mongers. There are three primary areas of job creation:

One is in building or programming robots, maintaining them, writing software, or teaching them skills, along with all the associated new jobs in supporting industry and infrastructure change. Many such jobs will be temporary, lasting a decade or so as machines gradually take over, but that transition period is extremely valuable and important. If anything, it will be a lengthy period of extra jobs and the biggest problem may well be filling those jobs, not widespread redundancy.

Secondly, AI and robots won’t always work direct with customers. Very often they will work via a human intermediary. A good example is in medicine. AI can make better diagnoses than a GP, and could be many times cheaper, but unless the patient is educated, and very disciplined and knowledgeable, it also needs a human with human skills to talk to a patient to make sure they put in correct information. How many times have you looked at an online medical diagnosis site and concluded you have every disease going? It is hard to be honest sometimes when you are free to interpret every possible symptom any way you want, much easier to want to be told that you have a special case of wonderful person syndrome. Having to explain to a nurse or technician what is wrong forces you to be more honest about it. They can ask you similar questions, but your answers will need to be moderated and sensible or you know they might challenge you and make you feel foolish. You will get a good diagnosis because the input data will be measured, normalized and scaled appropriately for the AI using it. When you call a call center and talk to a human, invariably they are already the front end of a massive AI system. Making that AI bigger and better won’t replace them, just mean that they can deal with your query better.

Thirdly, and I believe most importantly of all, AI and automation will remove many of the barriers that stop people being entrepreneurs. How many business ideas have you had and not bothered to implement because it was too much effort or cost or both for too uncertain a gain? 10? 100? 1000? Suppose you could just explain your idea to your home AI and it did it all for you. It checked the idea, made a model, worked out how to make it work or whether it was just a crap idea. It then explained to you what the options were and whether it would be likely to work, and how much you might earn from it, and how much you’d actually have to do personally and how much you could farm out to the cloud. Then AI checked all the costs and legal issues, did all the admin, raised the capital by explaining the idea and risks and costs to other AIs, did all the legal company setup, organised the logistics, insurance, supply chains, distribution chains, marketing, finance, personnel, ran the payroll and tax. All you’d have to do is some of the fun work that you wanted to do when you had the idea and it would find others or machines or AI to fill in the rest. In that sort of world, we’d all be entrepreneurs. I’d have a chain of tea shops and a fashion empire and a media empire and run an environmental consultancy and I’d be an artist and a designer and a composer and a genetic engineer and have a transport company and a construction empire. I don’t do any of that because I’m lazy and not at all entrepreneurial, and my ideas all ‘need work’ and the economy isn’t smooth and well run, and there are too many legal issues and regulations and it would all be boring as hell. If we automate it and make it run efficiently, and I could get as much AI assistance as I need or want at every stage, then there is nothing to stop me doing all of it. I’d create thousands of jobs, and so would many other people, and there would be more jobs than we have people to fill them, so we’d need to build even more AI and machines to fill the gaps caused by the sudden economic boom.

So why the doom? It isn’t justified. The bad news isn’t as bad as people make out, and the good news never gets a mention. Adding it together, AI will stimulate more jobs, create a bigger and a better economy, we’ll be doing far more with our lives and generally having a great time. The few people who will inevitably fall through the cracks could easily be financed by the far larger economy and the very generous welfare it can finance. We can all have the universal basic income as our safety net, but many of us will be very much wealthier and won’t need it.

 

Negativity has to end soon

There is much negativity about the future at the moment but post-Brexit, I am optimistic about our future. It looks better than it did before the referendum. Negativity spoils the short term but not the long term. Even if the next PM is rubbish – and once again it is a choice between bad and useless – they will be replaced by a better one in due course. Short term quite bad, long term very good.

The UK voted to leave the EU because most people thought we would be better leaving it, including me. Many commentators seek to project their own prejudices onto those who voted the other way, but all we know is that 52% voted to leave the EU; we don’t know why each one made that decision. Counting the votes told how many wanted to stay or leave the EU and only that; it can’t be mined to derive details of the mindset leading to that vote. It is not actually reasonable to infer that someone is a racist selfish moron from a leave vote. They might be, but you can’t derive that from the vote itself; you’d need other data. I’m not interested in debating all the various opinion polls about possible reasons, but it seem to me that there is little in common between typical poll stats for reasons given and the straw men put up by some remainers to demonise leavers. Demonisation just increases division.

Leavers voted leave for diverse baskets of reasons and attitudes, but the reasonable hope of probably most of them was that Britain could survive and prosper outside the EU, though economic reasons were usually secondary to escaping an undemocratic EU that was going the wrong way. Unfortunately we don’t currently have the competent and well-intentioned politicians and economic leaders we need to achieve the full potential so that needs fixed. We have some, but many seem to think it is more important to sulk in the corner, stomp their feet, do all they can to harm the economy, stir up division and encourage hatred, resentment and general bad feeling instead of putting their effort into making the most of a situation they didn’t want. They’d seemingly rather live in misery so they could say ‘I told you so’, but for some it might be that they actually don’t know how to do their jobs outside of the EU. They need to learn fast or be replaced.

That was to be expected, but it is a passing phase. Eventually, they’ll stop sulking and start doing the job they’re paid to do, or they’ll be replaced by others who can and will. Then the future freedom, prosperity and opportunity will come, slowly but surely, and it will become clear that the right decision was made and people will wonder what all the fuss was about. The UK will prosper, and the EU will either steer away from an embryonic Unites States of Europe towards a proper Common Market or evaporate via the domino effect, regardless of Tusk’s silly remark today that ‘there will be no sequel to Brexit’, which I suspect will go down as one of the worst predictions in political history.

Already, countries around the world are starting to discuss trade agreements with us. There is already enthusiasm for a new trading block of free and independent countries to deal with the EU. Currency and stocks and shares were always expected to take a short term hit during a period of uncertainty, and higher interest rates for out national debt was always expected to increase. Those are short term costs of a bigger long term benefit. Politicians, economists and bankers should work to accelerate the recovery and provide momentum to take us into improved prosperity. Nothing can be gained by moaning about how awful it is, exaggerating gloom or talking the economy down.

The media never accepts responsibility for anything, but they too play a part in the talking down and general negativity we’re seeing now. They don’t simply report things, they add huge negative spin and bias to make their own political points, to try to prevent Brexit or to reduce the changes and consequent benefits it could provide. They have also played a large role in the election of the new PM, strongly favouring one candidate (May) while undermining all the others. When the dust settles, it will be an excellent time to review the terms of existence of both the BBC and Channel 4 and especially to restore the impartiality the BBC once was famed for. Using the BBC or Channel 4 News as data sources at the moment is the intellectual equivalent of trying to survive by drinking swamp water full of decomposing sheep. Political bias in paid-for news channels is to be expected, but bias in organisations paid for equally by everyone is not acceptable, and if they can’t or won’t police themselves, then external policing should be imposed.

Some negativity was always expected after the referendum, whichever way it went. It must be just a short term hit, not become a way of life. We must not let it become the norm. Sulking benefits nobody.

 

 

 

 

 

Automation and the London tube strike

I was invited on the BBC’s Radio 4 Today Programme to discuss automation this morning, but on Radio 4, studio audio quality is a higher priority than content quality, while quality of life for me is a higher priority than radio exposure, and going into Ipswich greatly reduces my quality of life. We amicably agreed they should find someone else.

There will be more automation in the future. On one hand, if we could totally automate every single job right now, all the same work would be done, so the world would still have the same overall wealth, but then we’d all be idle so our newly free time could be used to improve quality of life, or lie on beaches enjoying ourselves. The problem with that isn’t the automation itself, it is mainly the deciding what else to do with our time and establishing a fair means of distributing the wealth so it doesn’t just stay with ‘the mill owners’. Automation will eventually require some tweaks of capitalism (I discuss this at length in my book Total Sustainability).

We can’t and shouldn’t automate every job. Some jobs are dull and boring or reduce the worker to too low a level of  dignity, and they should be automated as far as we can economically – that is, without creating a greater problem elsewhere. Some jobs provide people with a huge sense of fulfillment or pleasure, and we ought to keep them and create more like them. Most jobs are in between and their situation is rather more complex. Jobs give us something to do with our time. They provide us with social contact. They stop us hanging around on the streets picking fights, or finding ways to demean ourselves or others. They provide dignity, status, self-actualisation. They provide a convenient mechanism for wealth distribution. Some provide stimulation, or exercise, or supervision. All of these factors add to the value of jobs above the actual financial value add.

The London tube strike illustrates one key factor in the social decision on which jobs should be automated. The tube provides an essential service that affects a very large number of people and all their interests should be taken into account.

The impact of potential automation on individual workers in the tube system is certainly important and we shouldn’t ignore it. It would force many of them to find other jobs, albeit in an area with very low unemployment and generally high salaries. Others would have to change to another role within the tube system, perhaps giving assistance and advice to customers instead of pushing buttons on a ticket machine or moving a lever back and forward in a train cab. I find it hard to see how pushing buttons can offer the same dignity or human fulfillment as directly helping another person, so I would consider that sort of change positive, apart from any potential income drop and its onward consequences.

On the other hand, the cumulative impacts on all those other people affected are astronomically large. Many people would have struggled to get to work. Many wouldn’t have bothered. A few would suffer health consequences due to the extra struggle or stress. Perhaps a few small business on the edge of survival will have been killed. Some tourists won’t come back, a lot will spend less. A very large number of businesses and individuals will suffer significantly to let the tube staff make a not very valid protest.

The interests of a small number of people shouldn’t be ignored, but neither should the interests of a large number of people. If these jobs are automated, a few staff would suffer significantly, most would just move on to other jobs, but the future minor miseries caused to millions would be avoided.

Other jobs that should be automated are those where staff are give undue power or authority over others. Most of us will have had bad experiences of jobsworth staff, perhaps including ticketing staff, whose personal attitude is rather less than helpful and whose replacement by a machine would make the world a better place. A few people sadly seem to relish their power to make someone else’s life more difficult. I am pleased to see widespread automation of check-in at airports for that reason too. There were simply too many check-in assistants who gleefully stood in front of big notices saying that rudeness and abuse will not be tolerated from customers, while happily abusing their customers, creating maximum inconvenience and grief to their customers through a jobsworth attitude or couldn’t-care-less incompetence. Where people are in a position of power or authority, where a job offers the sort of opportunities for sadistic self-actualisation some people get by making other people’s lives worse, there is a strong case for automation to avoid the temptation to abuse that power or authority.

As artificial intelligence and robotics increase in scope and ability, many more jobs will be automated, but more often it will affect parts of jobs. Increasing productivity isn’t a bad thing, nor is up-skilling someone to do a more difficult and fulfilling job than they could otherwise manage. Some parts of any job are dull, and we won’t miss them, if they are replaced by more enjoyable activity. In many cases, simple mechanical or information processing tasks will be replaced by those involving people skills, emotional skills. By automating these bits where we are essentially doing machine work, high technology forces us to concentrate on being human. That is no bad thing.

While automation moves people away from repetitive,boring, dangerous, low dignity tasks, or those that give people too much opportunity to cause problems for others, I am all in favour. Those jobs together don’t add up to enough to cause major economic problems. We can find better work for those concerned.

We need to guard against automation going too far though. When jobs are automated faster than new equivalent or better jobs can be created, then we will have a problem. Not from the automation itself, but as a result of the unemployment, the unbalanced wealth distribution, and all the social problems that result from those. We need to automate sustainably.

Human + machine is better than human alone, but human alone is probably better than machine alone.

Technology Convergence – What’s your Plan? Guest post by Rohit Talwar

Rohit is CEO of Fastfuture and a long-standing friend as well as an excellent futurist. He and I used to do a joint newsletter, and we have started again. Rohit sends it out to his mailing list as a proper newletter and because I don’t use mailing lists, I guest post it here. I’ll post my bit immediately after this one. I’m especially impressed since his bit ticks almost as many filing category boxes as it uses words.

Here is Rohit’s piece:

Technology Convergence – What’s your Plan?

I have just returned from South Korea where I was delivering a keynote speech to a cross-industry forum on how to prepare for and benefit from the opportunities arising from industry convergence. South Korea has made a major strategic commitment starting with government and running through the economy to be a leader in exploiting the potential opportunities arising from the convergence of industries made possible by advances in a range of disciplines. These include information and communications technology, biological and genetic sciences, energy and environmental sciences, cognitive science, materials science and nanotechnology.  From environmental monitoring, smart cars, and intelligent grids through to adaptive bioengineered materials and clothing-embedded wearable sensor device that monitor our health on a continuous basis – the potential is vast.

What struck me about the situation in Korea was how the opportunity is being viewed as a central component of the long-term future of Korea’s economy and how this is manifested in practice. Alongside a national plan, a government sponsored association has been established to drive and facilitate cross-industry collaboration to achieve convergence. In addition to various government-led support initiatives, a range of conferences are being created to help every major sector of the economy understand, explore, act on and realise the potential arising out of convergence.

I am fortunate to get the opportunity to visit 20-25 countries a year across all six continents and get to study and see a lot of what is happening to create tomorrow’s economy. Whilst my perspective is by no means complete, I am not aware of any country where such a systematic and rigorous approach is being taken to driving industry convergence. Those who study Korea know that this approach is nothing new for them – long term research and strategic planning are acknowledged to have played a major role in the evolution of its knowledge economy and rise of Korea and its technology brands on the global stage. Coming from the UK, where it seems that long term thinking and national policy are now long lost relatives, I wonder why it is that so few countries are willing to or capable of taking such a strategic approach.

Rohit on the Road

In the next few months Rohit will delivering speeches in Oslo, Paris, Vilnius, Warsaw, Frankfurt, Helsinki, Denver, Las Vegas, Oman, Leeds and London. Topics to be covered include human enhancement, the future of professional services, the future of HR, transformational forces in business, global drivers of change, how smart businesses create the future, the future technology timeline, the future of travel and tourism, the future of airlines and airports and the future of education. If you would like to arrange a meeting with Rohit in one of these cities or are interested in arranging a presentation or workshop for your organisation, please contact rohit@fastfuture.com

Progress and The Care Economy (btw, the UN is badly wrong)

I’ve often written about the Care Economy, the one that I think comes after the information economy. As new things come over the horizon, it is always worth an update. And anyway, I promised a while back to write further on the future of capitalism: https://timeguide.wordpress.com/2012/01/04/we-need-to-rethink-capitalism/ so time to get on with it I guess. The Care Economy idea is resonating better with the way the word is now than when I first raised it in the 90s. We see a stronger desire to live sustainably, to see human skills valued per se rather than just financial wealth. These are both care economy values.

The primary driver for the care economy is progress in machines. Let’s include large-scale robotics and AI of course, but let’s also recognise that much of the progress now happens at invisibly small scales, in biotech, in synthetic biology, biomimetics, in synthetic neurology.  Taking the most obvious and most easily quantifiable area, the fastest supercomputers now compare to the human brain in overall power (which I estimate at the equivalent of around 10^15 instructions per second and 10^15 bits of storage, though it is a bit of an apples-and-oranges comparison). Thanks to the limits on Moore’s Law recently having been pushed back another decade or two, their descendants will carry on getting even better (graphene and molybdenene circuits can be smaller and faster, with lasagne processors not far away, not to mention smart yoghurt, so there is a lot of potential still in the pipeline, but that’s another blog). Eventually, even personal gadgets will have better capability than the Mk1 human brain (unless regulation intervenes).

An ordinary computer doesn’t work the same way as the brain of course, but work is also ongoing in understanding how the brain works, and scientists can produce electronic equivalents to some small brain regions already. Electronics isn’t all digital chips, there are many other sorts of devices too. With a big well-stocked toolbox and detailed instruction manuals, or descendants will be able to do a lot with electronics.

What then for your information economy job? Well, it will eventually be better, faster and cheaper to use some sort of machine instead of you. That will force you to retrain or to concentrate on those areas of your job that can’t still be done by machine, and those areas will be shrinking.

The Care Economy is recognition of this problem, and suggesting that we will focus more and more on the emotional, human interaction, side of work. Social, emotional, interpersonal skills will be relatively more important. Hence, for lack of a better name, the care economy. However, there is absolutely no guarantee that the number of care economy jobs will expand to fill the number leaving the information economy. Today, about 30% of jobs are in what could reasonably be described as the care economy. This can grow, but not indefinitely. So we will have to rework our economy to avoid excessive polarisation between haves and have nots. That won’t be easy. We will need to redesign capitalism.

It isn’t going to be just that a lot of people in information economy jobs will have migrated to care economy jobs. The nature of the economy will change. With machines increasingly doing the physical and intellectual work, it will be like a black box economy, where people put a request into the box, and out comes the required product. The cost of material goods will drop a great deal, as will the materials and energy needed – progress in all branches of science and engineering will accelerate a great deal as AI adds hugely to the available thinking. (Some of us call this the singularity, though that can be a somewhat misleading term, because infinite development speed is not possible.) A small number of people plus a lot of machine power will take basic resources (mined or recycled, it matters not) and add highly to their usefulness, vastly more than previous technology generations could. Nanotech, biotech, infotech and cognotech will converge and will allow tiny amounts of physical resource to yield huge benefits in people’s lives. NBIC convergence includes areas such as synthetic biology, biomimetics, which will adsorb parts of IT and strong AI as well as materials technology and nanotech. And vice versa.

I am not certain whether professional economists call it economic growth if we end up with far more stuff at lower output cost. Reduction in costs reduces prices, which reduces the size of the financial economy if growth in demand doesn’t grow faster. It is certainly a growth in the economy to me, since money is only one factor that indicates wealth and economics isn’t about money, it is about managing resources to gain the greatest benefit. And this benefit will grow spectacularly. In the care economy, we could even see less money but still all have a far higher standard of living. Money simply becomes less important as things become cheaper.

So one of a characteristics of the Care Economy is that it is a time of spectacular growth in material wealth, of plenty, even as it reduces environmental impact and improves the valuation of human interaction. Even if there is less of what we now call money (there may not be less money, I’m just saying it doesn’t necessarily matter if there is).

I find myself agreeing a bit, but mostly disagreeing with the UN’s recent proclamations here. (quick summary here:http://news.yahoo.com/un-panel-says-retool-world-economy-sustainability-164515165.html)

I fully agree that we need to become sustainable, and need to value non-financial things like quality of environment and human social well-being more. I believe strongly that the technology progress route is the best way to achieve it. The UN is very wrong with their approach. They are coming at it from totally the wrong angle, not understanding that technology progress can deliver lower environmental impact than cutting back on standard of living. Whether this is extreme left-wing influence or just bad futurist advice I don’t know. What is clear is that they argue for the opposite philosophy, that growth is bad, that we should trim back our lifestyles because only then can we live sustainably. That is nonsense, we don’t need to do that. In fact, to do so slows down the demand for new products slows down the progress to better ones that are more environmentally friendly. We are faced with a simple choice. Do we want to live in a healthy environment with happy people with a fantastic lifestyle? Or do we want a UN world of relative poverty, using primitive technology sparingly and telling ourselves it is for our own good, polishing our halos to make ourselves feel better?

The care economy will change our value sets as it progresses. If we leap towards the mature care economy, say 2050, where anyone can buy a $100 device with a five-figure IQ, and integrate it so well into their nervous system that it acts as a brain extension, what is the value of being smart? If anyone can use an assembler to create pretty much anything they can imagine (within modest size and resource limits), what is the value of physical skill? If anyone can use technology to reach what is today Olympic class performance in any sport within months, where is the value in being faster or stronger or more precise? Historical advantage has come from being born with a genetic advantage, and using cultural advantage to nurture it to overall benefit. Technology levels the field.

So we will value the most core of human skills, being human. Even if R2D2 can beat you in just about every way possible, it still won’t be human.

2050 is some way off, and the information economy is still running at full speed. However, we already see the increasing focus on human value and reduction of emphasis on financial wealth as indicators of happiness or even national well-being. We already see more demands for human value-add, such as ‘authenticity’, or provenance. Even celebrity is increasing in value. Some new trends will start soon. As people come to value machines less and humans more, companies will find the markets forcing them to become closer to the customer, to become more integrated into their customer communities. Many care economy businesses will emerge from social network sites.

The biggest problem with all of this, and it remains unresolved, is that increasing  efficiency via machine effort reduces the number of people needed in many job areas, and offers no guarantee elsewhere that new jobs will be created in equal measure. We don’t want to end up with many people unemployed and poor. We have to make sure somehow that everyone has access to the very nice life potentially on offer. We do need to redesign capitalism.

I wrote in my capitalism piece about taxing the accumulated human knowledge and infrastructure needed to make all the automated systems – those using them shouldn’t be able to keep all the wealth for themselves if the entire society has contributed, providing capital and effort is important and valuable, but nevertheless is only one of the inputs, and should be valued as such.

One idea that has started to gain ground since then is that of reducing the working week. It also has some merit. If there is enough work for 50 hours a week, it is perhaps better to have 2 people working 25 each than one working 50 and one unemployed, one rich and one poor. If more work becomes available, then they can both work longer again. This becomes more attractive still as automation brings the costs down so that the 25 hours provides enough to live well. It is one idea, and I am confident there will be more.

Concluding, we are one notch closer to the care economy. We can see a bit better where the technology path is leading, and can already see some of the signs of cultural change. We are also becoming more aware of some of the problems along the way, but are starting to produce potential solutions for them.  Sadly, we now have misguided institutions like the UN muddying the waters with policy suggestions that would destroy the potential for good, and make the world a worse place. The UN suggestions are based on poor thinking and bad futurology. They should be ignored.