Tomorrow is a big day for me. I’ve been invited to speak for a conference jointly organised by the Financial Times and Nanyang Technological University’s (NTU) Institute on Asian Consumer Insight.
I’ve been asked specifically to talk about philosophical issues related to artificial intelligence, robots, home automation, and other emerging technologies of the future.
I’m really excited, but I’m also very nervous because it’s a panel discussion with questions thrown at me. I’d be a lot less nervous if it were a talk, where I can prepare and plan in advance all that I want to say.
I can only anticipate what people will ask me. So, in this blog post, I’ll write all the things that I’ve prepared to say for tomorrow. I can only hope, with fingers crossed, that they will ask me questions along these lines. (This post is very raw, but I will come back to edit it after the event)
1. What is technology?
It’s easy to forget that technology is a tool that humans use as a means to fulfill a human purpose. It is designed by humans ultimately for humans, by exploiting either natural or social phenomena to achieve that function. There are two sides to technology: the “hardware,” referring to the physical things that will exploit the phenomena; and there’s the “software,” or the concept/logic that arranges and organises the “hardware” to fulfil that particular purpose. On the most basic level, it is the humans and our minds that function as the “software” of technology, manually controlling these tools to achieve what we want. On the more sophisticated level, it is the computer code that controls the computer system(s) to achieve a desired human effect.
On another level, we can understand technology as the collection of devices and practices that shape our culture. Technology is pervasive. It is present in our homes and in our work. It is what we use nowadays to get from place to place, and it is what we use to communicate with people. It is involved in giving us the food we eat, and the water we drink. Technology is everywhere, used in almost every single aspect of our lives to fulfil our human purposes. Technology is an essential component of human society and culture. We create the technologies that shape our culture. And it is this culture, which in turn, shapes the way we think, perceive, value, act, and respond to the people around us. Technology changes lives, for better or for worse, whether big or small. The very introduction of any piece of technology into a community will forever alter the path by which the community’s culture develops.
Technology is a tool which has the power, not only to help us achieve our needs and wants, but it has the power to shape our needs and wants, and how we understand ourselves and our role in the world. There is always a feedback loop between technology and humans.
2. Can technology save the world?
(I’m using the term, “save the world,” in particular because of the salvation narrative used to portray technology by some organisations or peoples. By saving the world, I’m referring to complex human/societal problems, or global challenges that confront nations from East to West, e.g. solving global warming, famine, political crisis, wars, etc.)
It’s interesting how many proponents of technology speak about technology as if technology (in the broadest sense of modern technology) can save the world, can solve some of the most pressing problems of the world, can make the world a better place for ourselves and for our future generations.
Some may cite the example of the atomic bomb as both a tool useful for international peace. It was what made Japan surrender, and it is what keeps the balance of power around the world. But some historians have pointed out that this is not true. It was other human factors, other human concerns, that led Japan to surrender. The Japanese were still more than ready to continue fighting even after the atomic bomb was dropped on them. This is one of many other examples of history where technology, no matter how great or horrifying it may be, does not save the world.
And of course, technology cannot and will not save the world. It is but a tool. And as tools, its effectiveness is dependent on the people using it. Tools are only instrumental to solving problems. Human problems, with all its complexities and complications, will remain human problems regardless of the amount of technology we throw at it. It would be naive to assume that technology – as a tool – will save the world.
But if we consider the impact technology has on culture, with its power to transform cultures, perceptions, thinking, and values, we may get a glimpse by which technology is the facilitator to “saving the world,’ or more accurately, in playing an instrumental and effective role for humans to resolve human/societal problems.
Let us consider the example of a bridge. We can build a bridge to connect one town to another town separated by a river. But in doing so, the bridge – like a catalyst – generates new means and opportunities for human interaction and for the exchange of ideas and cultures. The bridge is an instrumental means that becomes part of other human objectives. Over time, the interaction between the two towns will lead to transformations of their communities, transformations of their overall cultural outlook, ideas, production and economy.
Other influential technologies have the power to transform cultures as well. Of course, technology can transform culture for better or for worse, depending on how and what the technology facilitates and is instrumental for. But this is perhaps, for us, a clue by which we can understand the world-saving potential of technology – of its impact on culture as a whole, as an indirect means for achieving a “world-saving” effect.
3. Problems arising from technology are mainly human problems
One of the things we don’t expect from technology is that it generates more human problems than technical problems.
Why is it that more human problems come about? It goes back to the impact technology has on culture. As mentioned earlier, depending on how a piece of technology functions as an instrumental means for other broader human purposes, that technology can transform culture for better or for worse. Of course, this is seeing the situation too simply. The changes are better in some ways, and worse in other ways. Smartphones have created opportunities for us to interact with one another in so many wonderful ways, but it has also facilitated human laziness in so many other ways.
We need to recognise that many of the technological problems people are complaining about are actually human problems underlying these complaints. These are problems that will not be solved with more technology. A lazy person, for example, will continue to be lazy and exercise his laziness over all the technological tools in his possession (this I speak from experience). No amount of productivity tools will solve the problem.
The solutions are to be found in social, political and even ethical means. But perhaps part of the difficulty that we are facing now is that the rate of technological development is so fast, that our cultures are transforming faster than we can make sense of it, or to even identify the set of problems and solutions to them. This is perhaps something that we need to be aware of as a first step towards a more tangible solution.
4. Technology changes expectations
If we look at the history of technology, inventions like the cleaning appliances and computers promised to free up our time to pursue leisure or other meaningful activities. Instead, the complete opposite happened.
Appliances like the vacuum cleaner were supposed to reduce the time and effort required to clean the house. But it led to increased expectations of cleanliness. If you have a cleaning machine that cleans more effectively, how is it that your house is dirty? And with greater advances in cleaning technologies, the expectations continued to rise. What is interesting is that the concept of the housewife as one who looks after all the cooking and cleaning of the house, is a very modern conception born as a result of such cleaning appliances. Before that, women were working from their homes, involved in farming or textiles, while they tended to cooking, cleaning and child raising. But it was the increased expectation of a clean house, that made them so busy with cleaning, so busy trying to live up to the new expectations of a clean house, that they became too busy to work.
The same thing goes with computers. Before computers became the mainstream tool of productivity, they were marketed as a more efficient and productive means for work. You could save time working, so that you can devote more time for leisure or other meaningful activity. In fact, John M. Keynes predicted that in the future, we would only work 15 hours a week because technological advancements would have made our work easier. But all these didn’t happen. Why? Because our expectations of work had changed. If one employee could do the same amount of work in the less amount of time, it didn’t make sense for the employer to hire 3 employees. He could fire the other two, and let that one employee do the work of 3 people. And of course, if you could do the same work in a shorter time, you could also do the same work at a much higher quality in a short time too.
Machines are, of course, almost flawless in its operation and highly efficient, able to work long hours (or 24/7 even) without needing time to rest. That we use these systems so regularly at work and at home, it’s easy for such machine-thinking to leak into the way we perceive ourselves and others. Not only are we expecting people to do more work in the same amount of time, we have a tendency to demand that we work like machines.
This thinking is so prevalent that we find ourselves expressing it in our conversations from time to time. Here’s one:
“It’s so easy to forget that we’re not machines, that we need to rest.”
We see this machine-like requirement present sometimes in our hiring processes. We want people who are productive, efficient, least prone to error, etc. In short, we want someone as perfect as a machine!
It is interesting that future, emerging technologies are promising the same promises as the technologies before them – that we will be more productive and save time, that we will have more time for leisure and other meaningful activity. But history has shown time and again, that this is not the case because our expectations change, what we expect of ourselves and others have changed.
The real issue we need to consider is how AI, automation, etc., will change our expectations. Will it become more and more unrealistic? Has our society, perhaps, increased expectations more than our technology and people can currently support it? I’m saying this because in many big cities, and big organisations, the expectation to work long overtime hours has increased tremendously.
More importantly, we will humans expect ourselves to behave more and more like machines, and have less room for us to express our humanity? No room for error, for slowness, etc.? Are we creating a meritocracy based on machine-like perfection?
So, the issue we need to consider is: when we introduce new efficient and time-saving technologies, do we need to be aware of the way we market them? Is our marketing changing expectations faster than what is sustainable by technological progress and human capacity? Should we consider tampering expectations?
5. If all you have is a hammer, it is tempting to treat everything as nails
Abraham Maslow wrote:
“I suppose it is tempting, if the only tool you have is a hammer, to treat everything as if it were a nail.”
It is tempting to assume that every problem has a technological solution. And just as how hammering non-nails can be damaging to those non-nail issues, there are problems that arise from such an approach.
Part of the problem stems from the whole problem-solving approach. While it is useful for developing technologies to solve physical problems, it is trickier when it comes to human problems, to social problems. Human problems are complex and multi-dimensional. To solve the problem in a way that can be addressed by technology requires, first of all, that the problem be defined in a way that can suit a technological solution. This approach by reduces the complexity and richness of human problems. So while technology can be developed to addressed the defined problem, it ignores all related issues.
Here is an analogy to highlight another problem. A plumper is able to repair my toilet plumping because he has an idea/understanding of what a working plumping is. But what about societal/human issues? Can one develop a comprehensive idea/understanding of what the end solution is or should be? It is not possible. What we can envision is limited, and while we may develop a solution in that direction, it once again ignores everything else. This can and will lead to a lot of unforeseen consequences.
The problem of course, is that with this attitude of treating everything as nails, when things go wrong, the temptation is to invest more money on more technological solutions.
The underlying issue is this: Technology cannot save the world, nor can it solve all problems. It is a tool. The narrative we have about technology’s potential is highly problematic. Human problems must still be resolved by humans, by communities, and by a rich understanding of what it means to be human and the ways humans can flourish where they are.
We should reframe our problem-solving narrative to: humans can make a difference with the assistance of technology. It is not technology alone as if it has super miraculous powers, but humanity assisted by technology, humans using technology to bring out the best of other humans.
I think, it is essential to ponder on the complexities of humanity with close collaboration with the humanities and social sciences. This can and will lead us to richer understandings of what problems are, and how we can go about resolving them, and where applicable, with technology.
6. How far does the technological/robotic revolution have to go?
I will make a very provocative claim here, the point of which is to make you pause to ponder the extreme opposite view, so that we might find a balance in your own way.
Technological/robotic revolution can take a rest. There’s no need for it to go any further.
Bertrand Russell, after his long tenure of teaching in China, returned to the UK with a deep reflection about the problems of the West. He wrote – and this really resonates with me:
Our Western civilization is built upon assumptions, which, to a psychologist, are rationalizings of excessive energy. Our industrialism, our militarism, our love of progress, our missionary zeal, our imperialism, our passion for dominating and organizing, all spring from a superflux of the itch for activity. The creed of efficiency for its own sake, without regard for the ends to which it is directed, has become somewhat discredited in Europe since the war, which would have never taken place if the Western nations had been slightly more indolent. But in America this creed is still almost universally accepted; so it is in Japan, and so it is by the Bolsheviks, who have been aiming fundamentally at the Americanization of Russia. Russia, like China, may be described as an artist nation; but unlike China it has been governed, since the time of Peter the Great, by men who wished to introduce all the good and evil of the West. In former days, I might have had no doubt that such men were in the right. Some (though not many) of the Chinese returned students resemble them in the belief that Western push and hustle are the most desirable things on earth. I cannot now take this view. The evils produced in China by indolence seem to me far less disastrous, from the point of view of mankind at large, than those produced throughout the world by the domineering cocksureness of Europe and America. The Great War showed that something is wrong with our civilization; experience of Russia and China has made me believe that those countries can help to show us what it is that is wrong. The Chinese have discovered, and have practised for many centuries, a way of life which, if it could be adopted by all the world, would make all the world happy. We Europeans have not. Our way of life demands strife, exploitation, restless change, discontent and destruction. Efficiency directed to destruction can only end in annihilation, and it is to this consummation that our civilization is tending, if it cannot learn some of that wisdom for which it despises the East.
From a philosophical point of view, it is precisely because we have a linear conception of time and a linear narrative of progress in understanding and control of nature/universe that we assume that there should be a need for greater progress and development in our technologies.
Should things get better? Sure, why not. Should things be more convenient? Sure, why not. But why do we need things to get better, to be more convenient?
Why the discontent? Why do we not learn to accept things the way they are? I’m not saying this is what we should be doing, but I’m saying we should at least stop and ponder on this question.
Modern science and technology has given us the facade that we are in full control of our lives and destinies, and as long as we can arrange live in a certain way, we can achieve happiness. But it is interesting that this is a view that is fairly recent! But if you go back a few more centuries, you’d find that the philosophers of East and West have said that happiness/contentment can be achieved anytime, even now.
7. What is the potential of AI to replace jobs that we currently consider could never be done by a machine or an algorithm?
The way AI is progressing, I believe AI could very soon replace many low-level jobs.
But of course, the issue is whether companies are willing to invest huge sums of money on these AI systems. In some sectors, mass foreign labour is still cheaper than investing in new technologies that require far less manpower. There is just little or no incentive to switch over.
In this respect, the technology may be there, but there are other social/economic/political factors that would stand in the way of such adoption.
8. Who is responsible if a driverless car kills someone? What if an investment decision by a virtual financial adviser goes wrong?How can humans best adapt to ensure that machines are serving them and not the other way around?
One problem with the way the question is framed (“How can humans best adapt to ensure that machines are serving them and not the other way around?”) is that we speak of machines as if they have agency to control us. This in itself highlights a particular outlook that we have. It’s always so easy to push all responsibility to the machines.
In the recent Volkswagon robot accident that killed one man, the prevalent discourse was that the man was killed by a machine, the time has come where the machines are out to get us.
We can also talk about simple day-to-day activities. You try to do make a special arrangement with a particular organisation, and the first thing y0u hear is: “Sorry, I can’t do that, the system won’t allow it.”
Surprisingly, many of us are willing to accept this excuse, as if machines have full control. Or rather, it’s because we have difficulty taming these machines that we feel that the machines are in control.
It is precisely because we are so ready to give up all responsibility to the machines that we feel this way.
It is also this narrative that makes us feel that there isn’t anyone responsible if a driverless car kills someone, or if a virtual adviser gives the wrong financial advice.
What I want is to turn our attention to the developers. I’m not saying that we should hold them all accountable for everything.
Rather, these debates are problem today because of the way we have framed it. I think we need to have a more design-oriented, design-focused conception of safety and responsibility.
It’s not yet in our culture to develop responsible coding or responsible developing. I think what is essential is a paradigm of ethical design and ethical development, one that ensures not only that safety is given priority in development, but that the technologies are empowering. There are a good number of badly designed technologies that are so dehumanising, too focused on the function that it strips/robs the person of his/her humanity, and it leaves them feeling alienated or disenfranchised. And of course, in many ways, this leaves us feeling enslaved by technology because there isn’t much that we can do. It’s really about the design. Good design is humanising, and leaves people feeling empowered to embrace human goods. This includes robots too. We can design robots in ways that can be empowering and humanising to humans. It’s a question of whether or not we include these considerations into the design process, rather than focusing purely on function.
9. Is some of the Elon Musk, Stephen Hawking stuff about machines killing us etc., overdone?
For context, Elon Musk said:
“I don’t think anyone realizes how quickly artificial intelligence is advancing. Particularly if [the machine is] involved in recursive self-improvement . . . and its utility function is something that’s detrimental to humanity, then it will have a very bad effect. … If its [function] is just something like getting rid of e-mail spam and it determines the best way of getting rid of spam is getting rid of humans…
Stephen Hawking said:
“The development of full artificial intelligence could spell the end of the human race… It would take off on its own, and re-design itself at an ever increasing rate… Humans, who are limited by slow biological evolution, couldn’t compete, and would be superseded.”
A joke question that we should ponder is: If full self-improving AI technology is so scary, why are we even developing it? Why not spare the human race by not doing it?
I don’t have much to say in answer to this question. Here’s what I’ll say:
Of course, it is natural to fear what we do not know. After all, we will not be in control of self-improving AI technology, so we cannot predict what it’ll do to us, either directly or indirectly.
A few things underlying Musk’s view of AI (and that of many Hollywood movies). (1) One, is that humans are so bad that a more intelligent AI would need to eliminate us. And (2), that a more intelligent AI would see that we are a threat to its existence or to the survival of the planet, and thus must be eliminated.
I think a lot of these are projections of our insecurities. That someone or something better than us will take over and get rid of us. It is not necessarily the case, and it might be possible to forge friendships with them. Of course, some of us may prefer to see the whole thing as a power struggle. In which case, the Chinese perspective might be worthwhile: always keep your friends close, but always keep your enemies closer. A friendship and cooperation, even with the most intelligent being will always be worthwhile.
Hawking’s concern is more credible as he presents such AI beings as competition to our own evolution. If this is how AI materialises in the future, then it is a credible threat. Of course, this assumes that AI would compete with us and our niche for the same things, thereby competition would lead to our elimination.
But above and beyond all these, we really need to turn our attention to the design and development phase. That a large aspect missing from this is a concept of ethical/responsible development and design. Safety has never been a top priority in the history of inventions, until accidents occur. Perhaps it’s time we factor such considerations in our development stages.
10. How should we organise our working lives if lots of work we currently do is taken care of by machines? Will there always be new work created just as old work is destroyed, will we have to work shorter hours, or will it mean that some people work long hours (and are paid well) and others struggle to find work at all. In other words, will increasing mechanisation increase inequality?
I once attended a talk where the projection is that 15 years from now, if nothing changes, unemployment will be very high because the rate of technological development is so rapid, that people will not only lose their jobs because they are replaced by machines, but also because people do not have the time to learn new skills in such a short period of time to operate the new systems. Possibly the younger generation will have a better edge in learning these new systems much faster than us.
Of course, there are many other political, economic and social factors at play, that could prevent the widespread adoption of such automation, as evident in some sectors today, that still rely on mass labour because its still significantly cheaper.
But let’s assume that there is widespread adoption. As I mentioned earlier, expectations will increase, and so if history is a good gauge of what the future might be like, there will be new work created, but expectations of work will be greater than before. People will be expected to do the work of yet more people in a short period of time.