Can Data Science Replace Humanism and Solve the Problems of Strong AI, Free Will, and Self-Formation?

In order for data religion to replace humanism, it must solve the philosophical debates about the realization of strong artificial intelligence, free will, and self-formation, otherwise it will remain a mere paradigm.

 

According to Yuval Harari, religion is the reflection of superhuman laws in social structures and is an intersubjective reality that legitimizes human norms and values. Therefore, religion should not only provide a way of looking at the world, but also a standard for value judgments. An example that fulfills the first condition but does not fulfill the second is Einstein’s theory of relativity. Einstein argued that only the speed of light is absolute, and that the time and space in which all other objects exist is relative, depending on your point of view. This idea, backed by experimental evidence, has been accepted by many physicists, engineers, and even some laypeople. Similar to other religions, the theory has its critics. This is because relativity is known to be in theoretical contradiction with quantum mechanics, another paradigm of modern science. However, I wouldn’t call relativity a religion. This is because relativity does not provide a standard for value judgments.
Data religion provides a way of looking at the world, a paradigm, as Yuval Harari put it in Homo Deus. It’s about becoming more efficient at processing data. Life sciences have begun to view humans as data-processing systems, historiography understands history as a process of change in which data is processed efficiently, and political science describes political structures as mechanisms for collecting and analyzing information. However, the decades mentioned by Yuval Harari are unlikely to be enough for data science to displace humanism and become influential in value judgments, beyond providing a paradigm for viewing the world. This is due to the technological limitations of AI advancements and philosophical factors about free will.
One of the main reasons for the expectation that data religion will drive out humanism is that humans will become obsolete in the economic and political system. In the Data Church’s view, entities with better data processing capabilities will emerge that will outperform the human experience, which has brought about the best data processing capabilities to date, and thus will have greater value in the system, and humans will be devalued. This is because humanism places the highest value on the individual human being, whereas the data religion values people based on their contribution to data processing. This is happening now. Google predicts the spread of the flu faster than the U.S. health authorities, supercomputers analyze large numbers of observations faster than humans to predict tomorrow’s weather, and AlphaGo beat Lee Sedol by analyzing the checkerboard faster. In other words, from a data school perspective, humans are currently less valuable than machines when it comes to data processing.
However, this view overlooks an important fact. All of these machines need human users. Would DataKyo argue that a calculator or PowerPoint is more valuable because it processes data faster than a human? Probably not. It’s not the machines or programs that do the work, it’s the humans that use them. You might ask, how can you compare an electronic calculator or AlphaGo, but from the point of view of data processing power, they are no different in that they are both more efficient than humans in their respective fields (calculations, Go) and both require human users.
To better understand that there is not much difference between AI/machine learning and classical algorithms, we need to understand why these new technologies were developed. Before the development of machine learning techniques, one of the biggest questions in the computer science community was, “If computers can do so many things better than humans, why can’t they do it in the area of pattern recognition?” Many different methods were tried to solve this problem, and the most successful ones are what we now call artificial intelligence or machine learning. In other words, the only way these techniques differ from classical algorithms is that they are used in different domains. Sure, they work differently, but from the user’s point of view, they’re no more different than the gap between a calculator and a PowerPoint.
This principle only applies to weak AI. They can only be used in certain fields, and they don’t have a mind. However, a strong AI that has no rules and can learn a wide range of fields and create new things within them, even in the absence of a human user, would truly push back humanism and devalue humanity. So when will strong AI be realized? Fortunately, it is unlikely to be in the near future. The evidence for this is the very slow progress of strong AI to date. While the field of weak AI has recently achieved a high level of progress, the field of strong AI has seen almost no progress at all. This is because computers have very limited memory and processors compared to our brains. The brain has 100 billion neurons (nerve cells) and 100 trillion synapses (connections), which is far too many for a computer to directly mimic. Furthermore, our current lack of understanding of how the brain forms consciousness and has emotions makes the realization of strong AI even more difficult.
So, if it takes a long time for strong A.I. to be perfected, and the day comes when humans are completely defeated in data processing capabilities, will the Data Church be able to completely push back humanism? The realization of strong AI is just one of the many big mountains that the data religion has to climb. Another is the existence of free will. In his book, Yuval Harari explains that modern life science shows that free will does not exist, and he predicts the inevitable downfall of humanism because current liberal humanism relies so strongly on the existence of free will. He argues that data science is the idea that will step in to fill the void. But are these arguments enough?
The author deduces that free will does not exist by defining it as non-deterministic and non-random. But can we consider “free will” as something that can be “felt” as an inherent entity, like “pain” or “sadness”? In other words, can we define it as a subjective reality in the sense that we can feel its existence in the same way that we feel pain, sadness, etc. Even if the emotion of sadness can be produced by external devices and has no substance, only the firing of certain neurons, we do not say that sadness does not exist. Similarly, even if free will is scientifically denied or is merely a byproduct of mental operations, if an individual can feel it, we might define the emotion as free will.
Indeed, libertarian humanism seems to be based more strongly on free will as a subjective feeling than on scientifically refuted free will. In the case of theistic religions, the legitimacy of the religion is strongly dependent on the existence of God. If the existence of God is logically refuted, the whole point of religion is lost. Humanism is also an inter-subjective entity, meaning that it gains meaning from the stories people tell each other. However, unlike theistic religions, where followers believe in the existence of a deity and share that belief through the deity described in their scriptures (Bible, Koran, etc.), liberal humanism has its meaning through the shared existence of “free will” as experienced by individuals. In other words, even if an author explains the non-existence of free will through various conditions, it is not a logical basis for the collapse of liberal humanism.
The final mountain to climb for Dataism to truly become the new dominant religion is the formation of the self. The relationship between the self and strong AI is beyond the scope of this article, but it is worth recalling that no religion has ever become dominant without having a true self. All mainstream religions have an “I am” identity that binds them together and creates a strong social structure. Even religions that transcend the “I am” identity, such as Buddhism, have used that identity to exert social influence. Like free will, the formation of the self as a subjective feeling may be scientifically refutable. But, like free will, if it is actually felt, and if it provides a basis for identity, it will have meaning in people’s narratives. Can the current data religion bring about this kind of self-formation?
Therefore, in order for data religion to truly push back existing humanism, it is necessary to realize strong artificial intelligence, and furthermore, it is necessary to resolve philosophical and scientific debates about free will and self formation. If these conditions are not resolved, data religion will eventually become a mere paradigm.