The Great Machine War — and the fate of humanity

We often think of humans as a form of higher being, with a level of complexity that is untouchable by machines and other animals on Earth. However, studying Medicine and software engineering made me realise that stark similarities between human and machines.

Humans are not as special as I had always thought for the most of my life. Humans are machines with specific mechanisms improved and refined over millions of years of evolution. The human source code (ie. the human genome) was generated by randomly editing the genetic code and testing whether it improves survival rate. That’s a terribly inefficient method of engineering if you ask me.

Perhaps one might argue that it is the human brain that makes human so special. However, in recent days, I have noticed the increased similarity to between the memory and peripheral functions of the human brain and that of a computer.

In one instance, it seems that the human brain stores information in similar formats as that of a computer: arrays, key-value pairs and graphs. Beyond data storage, data retrieval occurs similar to that of a computer. It is easier for us to remember things when we index a particular set of data. For example, we can memorise a random list of US presidents and specific details of each president.

James Madison chief author of the Constitution, one of two presidents to sign it. President from 1809–1817.

John Adams was the author of the Massachusetts constitution. President from 1797–1801.

James Monroe studied law under Thomas Jefferson and established the Monroe Doctrine. President from 1817–1825.

Thomas Jefferson was co-author of the Declaration of Independence. President from 1801–1809.

However, if I was to ask you, which president served from in 1819? You would have to go through every single president you memorised and then recall the date, and then finally find the president with the date requested. This is similar to table scanning in a relational database, which is slow and inefficient. If instead, we memorised the list of US presidents by sorted dates,

1797 — John Adams was the author of the Massachusetts constitution

1801 — Thomas Jefferson was co-author of the Declaration of Independence.

1809 — James Madison chief author of the Constitution, one of two presidents to sign it.

1817 — James Monroe studied law under Thomas Jefferson and established the Monroe Doctrine.

Now, if I asked you which president served in 1819, you will be instantly be able to tell me the answer because the dates are indexed in your brain, and this process is faster and much more efficient. The implications of this is large. This means that depends on how we present information to students, the speed of recall and ease of memorisation can greatly be affected.

This is also the reason why many people cannot recite the letters of the alphabet backwards, because the data format stored in the brain is that of a linked list and the previous item is required to retrieve in the information of the next item.

Now, we could also remember the feeling of happiness and satisfaction of having a good meal. We would think that this feeling is something special. But really, this is simply a mechanism of chemoreceptors and stretch receptors sending specific electrical impulses to the brain. This is rather similar to your smartphone battery sensor sending a specific electric current to the CPU telling the CPU that the smartphone is charging. Our gastrointestinal system is a peripheral module of the brain, just as how the battery is a peripheral module of the smartphone.

With these examples, we realise that the design of the human body is nothing special. The human body is created by solving a complex set of engineering problems through millions of years of evolution. If humans could be created through randomly editing source code over million of years, it would then make sense that a machine with similar levels of complexity as a human could then be created by humans by deliberately writing source code in much less time.

If we were to consider why is humanity so dominant on Earth, one possible answer is that human intelligence has hit a critical threshold — the ability to create machines. The progress of human civilization is largely driven by technology, from the stone age, to the industrial revolution and now the information age. Technology is simply composed of machines that are created by humans. The ability to create functional machines enables humans to do less work but with greater output and this forms the critical foundation of innovation.

For this reason, the moment humans create a machine that is capable of creating other machines, humanity could then face an existential crisis as it’s position of dominance will be challenged. In this case, this machine would be able to evolve much faster than humans. This is primarily because the source code of this machine is created by a logical and deliberately designed programming language and written by intelligent humans. This makes the source code of the machine much more maintainable and hence it is easier for machine to edit and improve their own source code.

Humans, however, have no such benefit of a maintainable source code. The human genome is created by a brainless and randomly editing process called mutations and optimised through the process of evolution. The human source code is messy, filled with dead code (junk DNA) and basically as zero maintainability. This is one of the reasons why gene editing is exceedingly difficult, and it would mean that the rate of evolution via gene editing will pale in comparison to the rate of evolution of human-level machines.

With the great difference in the rate of evolution of humans vs human-level machines, we can expect then that these machines will very quickly exceed the intelligence of humans. These machines will then fight to quickly become the dominance force on the planet. This is what I will call The Great Machine War. For the first time in human history, the intelligence of the human species is being challenged by another species.

If we were to look around us, it is quickly apparent that the most intelligent species will become the owners and masters. Humans ride on horses. Humans have animals as pets. Humans use ‘dumb’ machines to do the hard work in mass production factories. Humans are superior as a machine and therefore rule over the less intelligent machines — computers, 3D printers, bicycles etc. It would then not be inconceivable that the machines of superior intelligence to humans will use humans as their slaves, to do the hard work.

This is a scary future to think about. But this is the future where we are headed.

There is still much suffering in the world. Healthcare is ridiculously expensive, poverty and the lack of access to basic goods and services are still incredibly prevalent in most parts of the world. However, to increase access to these basic goods and services, there is a need to increase the efficiency in production. This means automating more things with machines. However, just as in healthcare, we quickly realise that there is a limit to how much ‘dumb’ machines can do to help us. Eventually, we come to a realization that some tasks in healthcare are simply too complex to be done by dumb machines. This has spurred the shift towards ‘smart machines’ with the onslaught of machine learning and techniques using neural networks.

The shift towards smart machines is a double-edge sword. How smart must the smart machine be to be able to solve complex problems and reduce suffering in the world? At the same time, what is the minimum amount of intelligence required for a ‘smart machine’ to unlock the ability of creating other machines and henceforth start The Great Machine War? This is the great conundrum of the next generation. Will our attempts at reducing suffering in the world result in the complete annihilation of human dominance over the world?

It is then that we reach a rather depressing conclusion — the advancement of the human species will be limited by our own incompetence, our messy source code, the limits of our evolution. We will always be scared to create smarter machines because we are afraid they will be better than us and replace us. The utopia where all suffering can be alleviated is only but a distant dream, an unattainable dream owing to the limits of ourselves. The Darwinian tragedy pervades and the limits of the Lamarckian dream becomes all the more apparent.

The future of humanity lies in either one of stagnation of progress or a loss of dominance over the world.