Let us take a short break from the technical topics we have been discussing so far and start a conversation about technology from a philosophical point of view.
The technological advances that the world has seen in the past two centuries have brought enormous changes in human lives, society and economy and have transformed the world in numerous ways. Just think of how the industrial revolution, the computer technology and now the internet have shaped the environment in which we live in.
If we set aside all the negative aspects of this evolution, which were mainly caused by the way we have used the new techniques, technological progress has given us amazing machines that automate several processes. However, it would not be that much of an exaggeration, to say that it has only given us machines that assist us and tools for us to use. And tools still need an operator. The computer is the best example. An impressive machine, that can do tons of things for us, but can only do what we precisely tell it to do. We have to accurately describe the solution to a problem and it can easily fail if its data are not what we have predicted.
In spite of all this progress, there is still a goal that we have not reached, a goal on which more and more people have been working in the past few decades. That goal is true artificial intelligence, or to be more accurate with the terminology artificial general intelligence.
No matter how much our world has been transformed by the advances we have seen so far, nothing will be able to be compared to the impact of achieving real artificial intelligence.
I feel that the first step has been made and that is the setting of the destination. Larger and larger parts of the scientific community and the world leading IT companies are rigorously researching this matter. Even if there were exceptional researchers in the past working in that context, that paved the way, the attention it has attracted today is unprecedented. Advances in machine learning and cognitive science, applied in text mining, image and speech recognition, are examples of how far we have already gone.
But we are still very far from achieving artificial general intelligence. For example, driverless cars are already advertised as being around the corner, but I would like to respectfully disagree. I think that it will take time to replace a human driver in all traffic conditions. “All” is the keyword here. Just like with the cars example, we may have come up with solutions to several problems and conditions, but we have not achieved totality.
Naturally, a discussion as the one we have started here, sooner or later ends up being about Technological Singularity.
For those who are not familiar with the term, Technological Singularity, or simply Singularity, is a hypothetical point in time when machines will have surpassed human intelligence.
The most fascinating description I have seen, about what that moment would mean to life on earth is this: “All the change in the last million years will be superseded by the change in the next five minutes”, and comes from Kevin Kelly, founder of Wired magazine.
Surely that moment in time sounds spectacular, but why is it so important? Let us try to analyze it a little. Let us assume that time ts is the moment when Singularity is achieved. If at that point in time the level of human intelligence is A, the newly introduced machine will have intelligence A+, greater than A. This will mean that the new machine will carry all the intellectual properties of humans, including the one which allowed them to build a machine more intelligent than them. Therefore it will be able to build a machine more intelligent than itself with intelligence level A++, which will in turn will be able to build a machine with intelligence A+++ . A chain reaction will start and artificial intelligence will instantly reach levels that we will not be able to comprehend.
I am sure all these sound too hypothetical and much closer to science fiction than reality. But the question still stands: Is this achievable? Many brilliant people have given their opinion on the matter and if you search the web you will find interesting theories.
I will not even attempt to offer an opinion. But I will speculate on something else. Even if we may not be able to build a machine more intelligent than us, will we ever be able to replicate our own intelligence?
This is an entirely different question. I will unleash my imagination here and I will dare to say yes. That kind of intelligence is no science fiction. It exists in nature, it is us. It is only a matter of painstaking research and observation to be able to reverse engineer our intelligence and replicate it into the machines we will build. It is only a matter of time.
And the transforming power of such an achievement will be tremendous. We will no longer be making elaborate tools like we will have been doing until that day. We will be creating thinking machines, able to carry out all the tasks we are bothering ourselves with today (I am leaving all forms of art out of this, as I feel that that kind of creativity is somehow different from the kind of intelligence that we are discussing here).
Think of all the manual labour of today. All of it will be able to be done by the machines.
What about software? Will robots be able to fix software bugs? I think there will not even be the need for different kinds of software. There will instead be just one kind of software, able to do anything.
What about medicine and surgery? Will they be able to make diagnoses and perform surgical operations? Surely, yes. Perhaps even better than us. Imagine a doctor knowing everything that was ever said about medicine and using that knowledge. A surgeon operating with the precision of a machine, the intellect of a human and all the experience of all surgeons of all time.
Sounds wonderful? No! Do we suddenly somehow feel uncomfortable? Yes. Why?
Such radical changes will inevitably have indirect consequences and I am pretty sure that you are already thinking of that too as you read.
If we are replaced from all jobs what will happen to us? What will happen to society? Will we live happily ever after? Will we face judgment day? Unemployment and poverty?
We definitely cannot answer these questions with today’s experience. The established norms will just not work any more. Surely, as repeated many times before, the automatic elevator, left its manual operator unemployed and this was the fate of many other professionals. It was unfortunate and sad but replacing all professions is an entirely different thing.
Mankind will not just wait until it starves to death because no human will be needed to work. Social structure will change, economy will change. Everything will have to conform to this new structure. Currency will not be the same, if there will be any. Since, today we are mostly working for money, what will happen to money if we do not have to work any more? What about the goods and services we are purchasing with money? Will they be available to everyone? If not, how will they be distributed? Sociologists will have to be called in to provide some insight to these questions.
But even if we figure out how to live in such an environment and enjoy its fruits, there is still a dark side that waits to be explored. With all these great numbers of equally intelligent machines, able to replicate themselves and control so many functions vital to humanity, who will truly be in charge?
Think about this scenario. All food production is in the hands of the machines. They posses intelligence similar to ours. Can we trust that they will unquestionably follow orders and that their circuits will not develop free will or even worse turn against us?
We can have safety nets. We can have a universal switch that will turn them off if necessary and stop any upcoming crisis. But in such a case, who will produce food if we shut them down? Certainly, we will not be able to switch back to the old ways easily and wait until we fix the bugs. That will just not work. How will we work out problems like these?
The short answer is that I do not know. However, it is in my nature to be optimistic and I think that none of these will happen. We will conquer that technology one day, but we will progress gradually after that point. Achieving that level of artificial intelligence will mean that we will have first understood our own intelligence well enough, to replicate it to the level that we want and not beyond that.
And we might just stop at the point where robots are not humans, but only dream of becoming.
With that poetic phrase, I am stopping my futuristic brainstorming. Please feel free to comment, but keep in mind that all these are just thoughts that are not coming from an expert. Books can be written on the matter and a simple blog post cannot offer anything more than food for thought.
…It is strange. Coming back from this time travel and picking up my usual IT tasks suddenly feels like living in the stone age!
