Irreplaceable Us

I do not blame you for fearing that one day, your job too will be replaced by machines; and you are probably right. It will be. The question is, how far should we fear this replacement by machines? Would we all be rendered useless, replaced by stronger machines with far superior intelligence?

To answer this, let’s begin by turning the question on its head and list the abilities of humans that are lacking in machines:

The ability to make connections where there were none before

Can machines compete with humans at the upper end of human intelligence?

For machines to solve a problem, they require input in the form of data and code, which then form the parameters of the given solution. However, beyond this ‘pond’, machines are unable to make further connections. For a scientific problem for example, working within the parameters can only take you so far. Beyond that, you will need a completely different set of thinking.

Einstein, according to “Einstein’s Pathway to Special Relativity” by John D. Norton, “knew that there was something very right about Maxwell equation, but there was also something very wrong about it.” He continued,

“We have seen three components in Einstein’s discovery:

1. Astute analysis of new and surprising experiments

2. Deeply reflective philosophical analysis of the nature of time and physical theories

3. Solving an incongruous and overlooked problem in the foundations of electricity and magnetism.”

To advance, Einstein had to abandon the current way of thinking and formed his own path:

“Einstein had been reading many philosophers, including Hume and Mach. They had stressed that concepts are our servants, not our masters, and they are warranted only in so far as they might be grounded in experience. So was absolute simultaneity grounded properly in experience? Einstein began to think about the experiences that we use to establish simultaneity of events and he realized that it was not. Reading these philosophers gave him the courage to continue and to abandon absolute simultaneity. In its place came the relativity of simultaneity.” (1)

Dirac, on the other hand, discovered the relativistic equation for the electron. He also proposed the concept of a magnetic monopole, an object not yet known empirically. In fact, many other of his works increased in importance and significance as time goes by. Einstein kept a copy of Dirac’s book on quantum mechanics by his bed for reference. Even the scientists at the Large Hadron Collider frequently find that they need to refer to Dirac’s papers.

Dirac too, had to pursue other ways of thinking; as outlined in this conversation he had with Bohr:

Bohr: What are you working on?

Dirac: I am trying to get a relativistic theory of the electron.

Bohr: But Klein has already solved that problem.

Dirac: The transformation theory had become my darling. I was not interested in considering any theory which would not fit with my darling.

Thus, Dirac set out to find an alternative relativistic equation. (2)

“A great deal of my work is just playing with equations and seeing what they give.” Dirac

For new scientific discoveries such as the theory of relativity, the relativistic equation for the electron, the magnetic monopole and many others, the scientists needed to take a leap of imagination. This is what humans do best, to leap from one pond of parameters into another pond of parameters of a completely different application, and sometimes one that had no prior applications; then haul the catch back, and apply them to the original pond.

The ability to form conjectures

“We drown in data but thirst for knowledge” Associate Professor Jan Baumbach, University of Southern Denmark.

The ability to form conjectures is necessary when you don’t know where you are heading to, but you do feel the answer lies in a certain direction. Grok is the ability to grasp something intuitively. It is a word coined by Robert A. Heinlein for his 1961 science-fiction novel, “Stranger in a Strange Land”.  At other times, a part of us is instinctively aware of some conditions that have not yet been factored into our problem solving or decision making process.

In the words of Karl Popper:

“The belief that science proceeds from observation to theory is still so widely and so firmly held that my denial of it is often met with incredulity. I have even been suspected of being insincere- of denying what nobody in his senses would doubt. But in fact the belief that we can start with pure observation alone, without anything in the nature of a theory is absurd; as may be illustrated by the story of the man who dedicated his life to natural science, wrote down everything he could observe, and bequeathed his priceless collection of observations to the Royal Society to be used as evidence. This story should show us that though beetles may profitably be collected, observations may not.” (3)

The ability to gain inspiration

Humans gain inspiration from unlikely sources. For machines, once you input the problem and data, they will sit there crunching away. Are the machines going to go out later in the evening and attend an art show to view a Veronese painting, therefore giving them an idea to ditch solution path #2b14f for solution path #25xXCT instead?

Inspirations are important to investors finding investment themes, architects designing award-winning buildings or economists drawing new economic diagrams. The transfer from inspirations to concrete ideas is a process that is currently beyond the reach of machines. As long as abstract and unrelated ideas are illogically transferrable to concrete ideas, there is always a room for creative individuals in the workforce.

The possession of metacognition

Metacognition is the awareness and understanding of one’s own thought processes. What does understanding process mean to problem solving? It means that you will have the ability to quickly change your mind, to have an immediate understanding of the implications of new data. You will also understand that this then leads to the rise of new connections, new data format. Current AI development is progressing towards this ability, but attaining this ability that we take for granted is not easy to program into machines. For events that drastically and suddenly change a situation (i.e. real life), a machine cannot process this instantaneously; it needs new code and fresh data input.

Furthermore, can machines gather around in a meeting room and brainstorm a quick response to new circumstances and events?

At other times, they are things that are not clearly stated but which meanings are carried nonetheless through the words written. This ‘meta’ understanding is beyond the reach of machine intelligence, and I am not sure whether we shall ever be able to make Straussian-thinking Sufist machines.

Heuristics and Professor Gigerenzer

Heuristics refer to experience-based techniques for problem solving. Professor Gigerenzer asks how humans make inferences about their world with limited time and knowledge. In a world full of uncertainties, pile onto it a pressing need to make decisions quickly, humans usually need to use ‘rules of thumb’ to come up with a solution.

The question is then perhaps, how many machines have thumbs and if none, will machines be able to one day grow thumbs well enough as to have rules built around them?

The ability to understand that the wrong questions are being asked, put things into context, and to solve a complex problem simply

If you place a machine in a cart with no wheels in the garden and ask how it would transport itself across the lawn using the cart, it would come up with probably a dozen solutions as to how the cart can slide across the grass.

Place a human in a cart with no wheels; she would just walk out of the cart and get a different cart with wheels, or obtain a set of wheels and attach them to the cart. Either way, a departure from finding a literal answer to the given question is required.

The ability to tell the difference between a temporary solution and a permanent one

The fact that often times there are no perfect solutions and that sometimes, good enough for now is good enough. Kludge means to improvise or put together from an ill-assorted collection of parts; whereas Jugaad is a colloquial Hindi word that can mean an innovative fix or a simple work-around, sometimes pejoratively used for solutions that bend rules, or a resource that can be used as such, or a person who can solve a complicated issue. Kludge and Jugaad are such human words (in their raw necessity and desperation), that they should be used to partly define the nature of humans.

In solving a problem or carrying out a task, would machines cut corners? Is the inability to ‘cut corners’ a good thing or a bad thing? Slight shades of solution to a problem can make a world of difference. And this gradient of colourful reasoning and adaptation is what makes humans unique, because often solutions can come in the form of various guises.

Alternatively, can machines learn how to do things ‘elegantly’? Can machines do things with ‘finesse’ and ‘style’?

The ability to disobey commands

Is this important, you ask? Have you seen that submarine movie “Hunt for Red October”? Would you entrust the command of a nuclear weapon-carrying submarine to a machine? Or here’s a three-letter word: HAL.

The ability to pull the plug

For example, humans will not get stuck in an infinite loop, even when obsession grips them.

The ability to understand purpose and furthermore, to have ambitions

It is useful to think of machines as ships or vessels to get us across the sea, driverless though they may be, rather than thinking it is those ships that actually wish to get across the sea. We should pose too, the question, why would the ships want to get across the sea without the humans? What would the purpose be then? The answer is that it wouldn’t, it has no desire to get across the sea. We do.

Why then, did we build machines?

We first built machines to do the jobs we didn’t want to do, jobs that are back-breaking, or that machines can do faster or better. For example, we built the Spinning Jenny and Jethro Tull’s seed drill. Then, we built machines that enabled us to do what we wish we can do physically, but lacking the strength or ability to do so. The steam engine helped us with its strength. We wanted to fly, so we invented planes to help us travel longer distances in a shorter time, and access places that would have been otherwise hard to access.

We built machines that enabled us to do mind tasks that we wish to do, but couldn’t. Like calculating large sums in an instant, or regressing several factors at once and finding the mean and variance. Thus, we continued building machine after new machine that can progressively be ‘taught’ the things that we can’t do and to do the things we can do, better than we can hope to. Next would be to advance unsupervised learning for machines, a step that some AI companies have been thinking about for a while now.

Are there any tasks that are so far resistant to automation? Yes, the things that in order to advance to the next step require detachment from existing rules, previous trends, conventions and norms. Anything that requires the program writing the solution, to rewrite itself as the problem unfolds in real time.

Some AI companies are attempting to address these issues by using generative models and emphasising on network structures -basically, modelling the AI on human brains. They are attempting to reverse-engineer human learning and cognitive development as to produce more human-like machine learning systems(4). If machine intelligence is to advance to a very high degree, many of its qualities would be ones that are able to mimic human thinking. Put in another way, for AI to be of a high level use to us humans, there would be a lot of ‘human’ built into it.

After having replaced or supplemented humans’ strength and intelligence, which treasure in our human arsenal would be plundered next? Creativity?

This is already happening. I can see machines are beginning to be repurposed, to breach on the non-necessities, such as art. But until machines are able to appreciate art, are they able to produce meaningful art? Same goes with crafting poetry. Although from recent examples I saw on the internet, on computer composed poetry, I might already be proven wrong.

Data rich areas such as weather reporting or financial reporting are slowly being taken over by robot ‘journalists’. Yes, they can report in a concise and timely manner, but can they make a piece interesting, entertaining or moving? What about insights? Reading some articles written by some journalists today though, I feel that perhaps some of them are better replaced by robot journalists.

Measurement gauges will allow machines to produce products so perfect in creation that imperfection will come at a premium. Human artists would create vases that could otherwise be produced perfectly by 3D printers, the only difference is that they will be tiny nicks here and there and thumb prints baked into the clay. The vase would be ten times as much as the perfect 3D manufactured ones. That is, until someone figures out that they can upload a program purposely mimicking the imperfection of human produced vase and dwindle down that premium too.

So is the next step then, for us to teach machines the art of craft, or alternatively, to craft the arts? Or do we retain craftsmanship as a distinct identity of humans’ pursuit, away from the greedy metal claws of machines? To uphold craftsmanship and artistry as insignia of human creativity, never to be possessed by machines?

If one day, machines are capable of producing meaningful art that touches our souls, would we start losing our faith in our identity as humans, or would we be proud of humanity, to be able to produce such a creation that in turn, are able to create soul-moving art?

For the modern Luddite artists, there may be a need to go into isolation for fruitful creativity, or to have any creativity at all. Our world will be so crowded with transistors that the only place to get away from them will be in caves in a mountain on Tibet. These artists will produce such beautiful heart-wrenching arts that they become a new movement, not unlike Cubism and Fauvism.

Nevertheless, I believe that the heat of human creativity tempering the steel of algorithmic rules and randomness has great promise to bring an exciting revolution in the arts world.

Moving forward, there is a temptation to replace every imaginable, possible human task with machine input. This is good to a certain extent; but is this fully desirable? A metaphor would be to replace every single meal with Soylent because it is economical, convenient and nourishing. What would the consequences be?

Imagine if every single share purchased at the exchanges was as a result of computer algorithm. Would the financial markets be extremely efficient in the absence of human decisions and their behavioural fallacies? There would be no central banks setting rates. Instead, all economic rates, be they interest rates, exchange rates or taxes are determined by an international ‘bank’ of servers. Would we then, still have a need for economic professors to explain the economy to us?

The winners would be successful because they make the machines, understand the machines, use the machines productively, or even use the machines to make other machines. There should be a novel of a human world where everything is automated (even feeding oneself, not unlike the fat captain in Wall-E) and all the decisions are entrusted to the machines. The novel would make a good movie too. I think we should all re-watch Wall-E by the way, an underrated movie of what the future could be like.


Teaching how to code as a subject in schools

People will dismiss this idea of teaching everyone to code, saying it’s not a matter of providing the education, but rather a large number of population are not interested enough to code, or coding is too complex to learn.

Remember two centuries ago when many people couldn’t read and write? Did humankind as a whole say, never mind, those who were interested enough to learn how to read, smart enough to be able to read, may attain knowledge, and benefit from the literature available? No, there was a concerted effort in increasing the world’s literacy, and as a result, today’s global literacy rate for all people aged 15 and over is 84.1%. That is because we understood that humankind’s intelligence as a whole does rise over time, that education and knowledge is important, and that everyone should have access to knowledge through reading.

It is the same argument that I put forward, that we must realize just knowing how to read is no longer enough, that the ability to speak the language of machines is an essential skill for everyone to have, at least until machines reach the level where we can program using ‘human’ language. Who knows, a hybrid language may emerge from this process where the vocabulary’s etymology is a quarter from Mandarin, another quarter Hindi, and the other half from C++.

Setting aside negative perception or fear of machines

In the distant future, I hope with the help of machines, we will be released of our daily, boring need to spend our time on work that sustains our bodies, and concentrate instead on tasks that sustain our souls.

Make it a nation’s policy to ease access to technology for everyone, instill in us the means so that we may become a nation of tinkerers, subsidising 3D printers so that they can access whatever tools or information they wish on-line to make the printer work for whatever they want it to work for. This also nurtures a rich environment for open-source efforts. In fact, there should be a policy to open-source everything that can be open-sourced. Tesla recently, has made its patents open to all, a move which hopefully will spur future advancement in electric cars and batteries.

Perhaps as a jump-start, we should ban the production and import of plastic forks, and anyone wishing to hold a party or a picnic must learn themselves how to 3D print plastic forks.

Shifting the focus

According to estimates by Frey and Osborne, about 47 percent of total US employment is at risk from computerisation (5). Furthermore, according to the study by Beaudry, Green and Sands (6), in the year 2000, the demand for cognitive tasks associated with high educational skill in the US job market underwent a reversal.

Maybe the focus shouldn’t be on what kind of jobs will be displaced by machines, but rather what new jobs can be created with the help of machines. In 1956, how many would have thought that millions will have a job description as ‘software developer’? Therefore today, what bizarre, a leap of imagination, job description can you conjure up with the knowledge that we are not inhibited by our brain processing power or bodily strength?

Can we repurpose ourselves if machines take over our original purpose? How would we repurpose ourselves?

This is potentially the most difficult question for humanity to resolve. On the one hand, Buckminster Fuller painted a very optimistic picture of the future of work: “We should do away with the absolutely specious notion that everybody has to earn a living. It is a fact today that one in ten thousand of us can make a technological breakthrough capable of supporting all the rest.” On the other hand humans need to have a sense of purpose: studies show that old people die earlier if they lack such sense in their life.

Intelligence Amplification and Argument Mapping Software

This is also known as ‘machine augmented intelligence’ where information technology is used in augmenting human intelligence. The possibility of every human having the capability of a savant has effects that are far-reaching. In some ways, Googling has afforded us the memory and information of almost every subject in the time that it takes us to type the search terms.

Argument mapping is a visual representation of the structure of an argument. There is software that allows you to map your arguments, therefore giving you a tool to explore every logical nuance of an argument. It also flags certain isolated nodes of your argument where it is disconnected from your main points or where it should not even be included.

Intelligence amplification (IA) allows us to have better quality information to form our conclusions, whereas argument mapping software can instill the habit of rational thinking. Once these tools are widely adopted and effectively employed, they have the ability to change how society thinks, and therefore, how we vote and our appreciation of long term aims. Issues such as the environment or minority rights would perhaps have greater priority in policy making as politicians are afforded better decisions knowing their constituents have the correct priorities in mind and better understanding of why certain policies are beneficial and some not.

A well-informed, highly rational society – where would that take us?

Dynamic Integration, a Symbiosis and Verschmelzen

The era of big data makes it even more important that the human interface is designed as a process imbued with as little friction as possible.

More and more machines are designed with human usability, thinking and creativity in mind, and in turn, we really should educate our children with machine capabilities in mind. This dynamic integration of ‘machines for humans’ and ‘humans with machines’ would bring about a useful symbiosis where humans will be able to create better and better machines; and machines may help humans attempt to reach their full potential.

“Man-computer symbiosis is a subclass of man-machine systems. There are many man-machine systems. At present, however, there are no man-computer symbiosis. The purposes of this paper are to present the concept and, hopefully, to foster the development of man-computer symbiosis by analyzing some problems of interaction between men and computing machines, calling attention to applicable principles of man-machine engineering, and pointing out a few questions to which research answers are needed. The hope is that, in not too many years, human brains and computing machines will be coupled together very tightly, and that the resulting partnership will think as no human brain has ever thought and process data in a way not approached by the information-handling machines we know today.” (7)

Verschmelzen is a German word which means to merge, fuse, blend, coalesce or unify. The development of machines and other technology should be in parity with that of humans’; but to jump-start this virtuous circle, certain steps must be taken quickly and aggressively. In fact, the faster humans adapt to new technology, the lower the level of future unemployment from mechanisation and computerisation could be.


Perhaps alongside productivity, we should start measuring average free quality time per capita (FreeQuTiP). This measures how much free time you possess to pursue activities not pertaining to livelihood, free time to take that belly dancing class that you’ve always wanted, or build that toy plastic model of Starship Enterprise. To involve yourself in charitable activities or to write that doorstop novel you’ve always wanted to. Further integration of machines into the society will increase the level of FreeQuTiP.

The real difficulty with the ‘rise of the machines’ is that those who don’t adapt, even though benefiting from machines, would feel that they are not in charge of their own lives. They would feel that ‘true’ success or wealth would be inaccessible because of their inability to utilise and benefit from machines as successfully as those who do. This might be damaging to the self-esteem of those who cannot compete. Their feelings, even with elevated welfare, would be that they are receiving handouts. The key to human happiness is to have the ability to earn their rewards rather than the rewards being freely given. A whole society equipped with personal iPads but have nothing to do on those gadgets short of the usual entertainment would not be a happy society.

The Future

I began this piece by asking what makes us humans unique, or even superior to machines – and came up with a few examples. I then ask, would these qualities one day, be replicated and therefore usurped by machines? Unsurprisingly, the answer is yes, and many humans are working towards making that a reality. Well, if our current irreplaceable qualities are replaceable in the future by machines, how should we view this development?

There are only two options that I see, halt the progress of machines, or encourage it. Surely, there are no remaining Luddites in today’s world – we like our iPads too much, and therefore, we must encourage the progress. But is that enough?

The answer is no, it is not only machines’ progress that we must encourage, but humans’ progress in parity and in anticipation of machines roles in our lives – that is the secret sauce. If you love going around in your car, you must take driving lessons, if you want to benefit from AI and other related machine technologies, then you must learn how to fully utilise them. Painful for the current generations, but we must put these skills and knowledge in place for the future generations, who I believe will take this learning in their stride.

In Malaysia, during World War II when the Japanese invaded, local people were left without food. They had to dig the ground for Cassava, which is like a potato that grows wild underground. They survived by boiling and frying the Cassava. When they got sick of eating Cassava in its plain form, they ground it up into flour and made delicious desserts out of this plain and ugly root plant. When the low hanging fruits are gone, it may be time for us to start digging deep and believe in our own uniqueness. To truly understand that that very root of our humanity, those abilities that we naturally possess, in tandem with the rise of the second machine age, will continue to nourish and sustain us.

As we witnessed in the first machine age, humanity should look forward to a period where we achieve our full human potential by partnering with machines, and not by resisting them. “Resistance is futile” as the Borg said; but this human would rather say, “Partnering is progress”.





(4) “The Future of Employment: How Susceptible Are Jobs to Computerisation?”, Carl B Frey, Michael A Osborne, September 2013.

(5) “The great reversal in the demand for skill and cognitive tasks”, Paul Beaudry,David A. Green,Ben Sand January 2013.

(6) “Conjectures and Refutations: The Growth of Scientific Knowledge”, Karl Popper, 1963.

(7) “Man-Computer Symbiosis”, Licklider, J.C.R., IRE Transactions on Human Factors in Electronics, vol. HFE-1, 4-11, Mar 1960. Eprint.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s