😙 ♒️ 👉🏼 “Criticism of pure reason” for programmers 🚢 🤦🏼 ↖️

Everyone who for some reason read or tried to read the “Critique of Pure Reason” by Immanuel Kant, noted that this work can hardly be called easy to read. Wading through complex sentences half a page long, it is not easy to catch the spirit of his system, and indeed often understand what is at stake. What is noteworthy, Kant himself perfectly understood the literary flaws of his work, but in an effort to systematically expound the result of many years of reflection, as he writes: " ... I had to omit or shorten much that is not essential for the completeness of the whole, although it might be desirable for some readers . " This article is an attempt to outline an analogy connected with the principles of computer systems, which helps me to connect Kant's concepts.

To understand the motives for writing criticism, you need to take a short digression into the history of philosophical discussions about the possibilities of human knowledge. The question of what truth is and how much the human mind can claim to know it has excited the best human minds since ancient times. The original antique systems, firstly, introduced an important distinction between the concepts of “truth” and “opinion”, and also laid the foundations for such a methodological approach as analytics.

During the Middle Ages, Aristotle's analytics was significantly developed as part of the justification of scholastic theological concepts. An attempt to use the human mind to prove the existence of God led to the fact that everything that was "illogical" became "dogmatic."

The philosophy of the New Age brought a wave of faith in the human mind and it became true that it was rational, if it is logical this is the truth. But in the wake of the success of using this approach in science, again attempts have been made to rationally justify super-cognizable concepts: the existence of God, the immortality of the soul, free will. The problem of conceptual deductive systems proclaiming eternal truths based on "evidence for reason", in the use of seemingly unshakable concepts like causal relationships.

The famous “from X it is easy to see that Y ” is very susceptible to substitution of X and Y, which are invalid for this law, into it. It is possible to verify the proposition “force of action is equal to force of opposition” in practice, and the statement “if there is a reason for all the consequences, then there must be a root cause” no longer exists.

Experimental verification of judgments became the basis for another trend of philosophical thought - empiricism. The absolutization of experience and the diminution of the role of universal theories and concepts led to doubt in induction as an approach to the construction of laws. Most clearly about the problem of correlation of individual experience with universal laws, the British philosopher David Hume said. In fact, the habit of associating phenomenon A with phenomenon B (objects fall down , swans are only white ) does not mean that it is “ necessaryis the only way "and" vseobschno all swans and things. "

Kant said that such criticism of causality awakened him from dogmatic slumber and forced to speak in defense of cognitive abilities of mankind, in the ability to generate knowledge that will be necessary and universal .

For of understanding the path that Kant has traveled in his work, I propose to introduce an analogy and consider a Computing System ( BC ) that records signals from the External World ( VM ) using a Data Receiver ( PD ) for processing in the Block Data Processing ( AML ) and entering into the Knowledge Base ( KB ).

In terms of critical philosophy, this is respectively:

Computing system - the human apparatus of thinking
The data receiver is the sensory apparatus, our five senses
Data Processing Unit - Mind
Knowledge Base - accumulated knowledge
The outside world is the world of things in itself

We will present our interaction with the outside world, a single act of cognition in the form of a certain data transaction and its processing in our computer system. The general task with this approach is to explain how it is possible to fill our knowledge base with true judgments, and now it becomes the truth, which is true for each such transaction.

According to Kant, the filling of the Knowledge Base begins during interaction with the External World and follows it, but not only the VM is the source of such knowledge, the features of our armed forces also affect the contents of the knowledge base .

The idea is quite simple, despite the fact that Computing Systems are different, if you select those principles that are common for building all VS, you can use this as an invariant for any possible transaction.

What is so universal and necessary in our system ?

Let's see how you can organize the components of our computing system.
Our data receiver is a block (not quite an ADC, but such an analogy will come down in our circuit), which provides the processed information in a kind of understandable form:

We do not know anything about a specific set of signal bits, but we know for sure that they are somehow “positioned” and act as changing data sets in time. The interpretation schemes for such data (bit depth, processing frequency) can be different, but in the end we should get the characteristics of this information - the number of bits in a certain group, their presence in general, patterns in consecutive messages, etc.

Turning this streaming dataset into a complete picture is the basis for our primary task of filling or updating our knowledge base. Since a simple thread does not give any holistic picture, at some point we should start associating the current set with a number of previous ones.

This basic ability of our system is ensured by the ability of the Data Processing Unit to save the state of the system, in a very rough representation, performing such a simple algorithmic operation:

if (previousContext.hasDependency(current)) { 
    updatePreviousContext();
}

In Kant's terminology, this basic ability to “maintain a state of consciousness” regarding sensory data is called the unity of apperception (conscious perception) and is the basis for all subsequent rational judgments.

Without this conservation of “I think” in every act of perception, the synthesis of knowledge itself is impossible, it is the basis for every possible transaction with the outside world.

The transformation of sensations (data) into concepts (data types) and, finally, into foundations (software constructs) is the chain that is based on the transcendental unity of apperception.

It is important to understand that this principle of binding cannot be applied in isolation from transactions, from interaction with the outside world, from possible experience.

That which is the basis of knowledge of the world of phenomena cannot be freely applied to super-experienced concepts (God, freedom, soul).

But it’s very difficult for the mind to limit this “flight of consciousness” and a person often very simply crosses the border of his “computational abilities” and begins to “program concepts” without reference to the Data Receiver, based on the accumulated data processing concepts ( categories from Kant).
Such judgments that arise without being tied to the Outer World by Kant are called a priori , and the question is whether it is possible to fill (not just interpret existing records in the KB, analytically, based on the law of contradiction), but expand ( synthesize) the contents of our Knowledge Base without using data from the outside world is key for him - how are synthetic a priori judgments possible?

Kant says that since we know that the principles of the Data Receiver work and they are always applied for each transaction, then for each subsequent one they will be fair.

If we know that the Data Receiver has spatially placed (bit positions on the grid) and temporal (bit changes in each cell) characteristics, we can create laws based on this characteristic of this data.

Kant believed in the truth of mathematics, and its geometric spatial constructions that use abstract space and abstract construction procedures (a line must be drawn in order to be intuitively understood as a dimensionless infinite entity). Such a priori constructions, in his opinion, are true precisely because they do not use more than the “basis of every possible transaction” with the world.

Even if we don’t know “how it really is”, the limitations of the Data Receiver, our sensuality will work for us each time, and this is the common thing between us and the External world, the properties of which make it possible to produce such desired synthetic a priori judgments , universal and necessary .

If for pure mathematics the properties of the Data Receiver are sufficient (the space for constructing geometry and time for calculation in arithmetic), then natural science already requires algorithmic support, at least a causal relationship as a basis for constructing judgments, such as the laws of conservation of energy. Proceeding from the same need to use the unity of apperception for each act of cognition, Kant uses a categorical apparatus (quantity, quality, causality, possibility) to explain the possibility of nature sciences.
This causal relationship may not be an objective state of affairs in the world (who knows whether there were any interruptions in the signal, whether everything was processed by the Data Receiver), but rather the subjective property of our algorithmic unit, without which the procedure for creating knowledge is not possible.

Problems with the use of a priori forms of sensuality (space and time) and a priori forms of reason (categories of quantity, quality, attitude and modality) begin where they are used in isolation from a possible transaction ( possible experience ) with the External World.

For example, our “internal programmer” has a burning desire to explore on the basis of “logic” and “feelings” his internal structure, the way he organizes his Computing System.

Since our “event processing cycle” seems to us, firstly, to be single, and secondly, to be absolutely real, we are trying to attribute the same structure to the Sun and other entities. Moreover, we are not limited to the time of the computing session (while Sunhave not been unplugged) and we think that even beyond the limits of possible experience (which ends with the death of the system), our aircraft can continue to exist. Unfortunately, such paralogisms of the mind in the context of judgments about the immortality and substantiality of the soul go beyond the scope of “possible transaction” and possible experience. Our computing system is the basis for transaction processing and nothing more, it is as unrecognizable according to Kant as the outside world (in fact, the “program does not know” the details of its runtime, in the cloud, distributed or somehow).

As for the outside world, then if we start to think in terms of an infinite “tape” of AP or infinite computational abilities of AMLwe can get equally logical judgments ( antinomies ), for example:

“The world is finite, the world is infinite” (theoretically, the PD can have arbitrary bit depth)
“The world is predictable, the world is random” (we simply cannot calculate this, or there is fundamental freedom)
“There is a root cause of everything, there is no such root cause” (something equally computationally powerful and circumventing the limitations of our PD ).

According to Kant, the idea of such absolute (in size, duration, or “power”) concepts does not in itself serve as a source of data, it only allows you to move in the direction of expanding existing knowledge.

Summing up, we can say that such a critical study of the capabilities of the "computing system" of a person gave Kant the basis to protect the already accumulated mathematical and natural science knowledge and to warn against the use of reason in super-experimental cases.

“Criticism of pure reason” for programmers

More articles: