The Problem and Idea of the Ethic Calculator

Many of the answers people give to ethical questions are influenced by culture, time, religion, habit, emotions, and personal interests. The question is: Can ethics be examined in a way that does not make its conclusions dependent on personal feelings, interests, and habits?

To understand the idea behind EC, think of an ordinary calculator. A calculator has no feelings, preferences, or personal opinions. It simply operates according to logical rules and produces correct results.

The Ethic Calculator, or EC, is not intended to be an entity with feelings or personal preferences either. It is intended to be a method for examining ethical arguments.

The foundation of EC’s method is reason and reasoning. EC is designed for someone who accepts reason as a tool of judgment. Therefore, it does not begin with a predefined moral principle or criterion. By reason here, I mean a cognitive faculty that, in addition to the ability to reason logically, is also capable of perceiving pain, suffering, and pleasure through the analysis of linguistic information derived from human senses.

The starting point is simply a linguistic proposition or argument. The first question is:

Is the argument sound?

EC examines the structure of an argument: its propositions, premises, the relationships between them, and whether it is logically possible to move from one proposition to the next.

Every reason introduced into this process must itself be examined.

Furthermore, even when EC encounters a perspectivist or consequentialist system, it examines the strength of that system’s arguments from within the framework of the system itself.

For this reason, EC resembles a “dry philosopher”: a philosopher whose judgment is not influenced by personal feelings, preferences, or human habits, and who follows only the structure of the argument.

Following Arguments to Their Roots

Suppose we say, “A is correct because of B.” Then we ask: Why B? Because of C.

The path therefore looks like this:

Action or judgment ← Reasons and premises ← Foundation ← Deeper foundation

At every stage, the central question is simple:

“On what basis is this proposition accepted?”

To visualize this structure, imagine a tree.

The trunk represents the action or proposition. The roots represent its reasons and foundations. The branches and leaves represent its consequences.

The farther we move from the trunk toward the roots, the finer and thinner the roots become, eventually resembling tiny capillaries. This is because, at that level, we may encounter very precise, subtle, and complex logical relationships.

However, an argument does not always follow a straight line. Sometimes a proposition requires several foundations to be accepted. EC must examine the different roots and, at times, consider several foundations simultaneously.

At every stage, the same question is repeated:

“Is another foundation needed to accept this proposition?”

If not, we have reached a fundamental basis within the structure being examined.

If another foundation is needed and we find it, the examination continues.

If another foundation is needed but cannot be found, the root remains incomplete.

And if the foundation that is found is itself illogical, that line of reasoning comes to a halt.

“Fundamental” here does not mean the ultimate truth. It simply means that, within the structure being examined, no further foundation has been deemed necessary to accept that proposition.

For example, this method can be applied to the examination of an argument concerning the torture of an innocent person. In this argument, the increase of pleasure is presented as the perspectival basis for justifying the action. However, the pleasure of the person being tortured has not been taken into account. The chain of reasoning is therefore incomplete, because one of the relevant aspects of the matter—the pleasure of the person being tortured—has been left out of the examination. Until this deficiency is addressed and a rational argument is provided to justify the action, performing it cannot be considered rational on the part of the agent.

Furthermore, failing to find a reason does not mean that no reason exists. Another reason may exist that has not yet been found or cannot currently be examined.

Therefore, EC does not need to prove that an action is wrong under all circumstances and for all time. It may simply show that the reason offered for the action is insufficient at an essential point in the chain of reasoning. And, for the time being, performing that action is not rational.

Furthermore, the existence of initial value criteria in perspectival systems does not obstruct EC. If an action can reach the system’s initial criterion through a logically coherent chain of reasoning, starting from the action itself, and that chain is complete, then the action is rational. Conversely, EC can expose weaknesses or logical errors in the chains of reasoning offered by those defending an action, without needing to evaluate the system’s criteria.

The distinction between the validity and soundness of an argument is well established in logic: the logical validity of an argument alone does not guarantee the truth of its premises. However, EC demonstrates the opposite of this view and makes it possible to evaluate the ethical status of an action independently of its premises. In other words, the problem lies in the philosopher’s lack of insight or failure to examine closely enough to expose the weakness in the action’s chain of reasoning, sometimes resorting instead to perspectivism and the initial criterion. In philosophical terms, an error in reasoning consists of jumping to conclusions based on an unjustified assumption due to a failure to adequately consider alternative lines of reasoning.

The Agent’s Intention

EC does not use a person’s intention as the basis for judgment.

Someone might say, “I had good intentions.” But rather than examining the intention, EC examines the reasoning itself:

Is there sufficient rational justification for doing this or not?

The Consequences of an Action

EC does not use the consequences of an action as the primary basis for evaluating the action itself either.

“The consequences of an action, considered as a complete set across the entirety of time and space, cannot be fully determined or traced. Therefore, consequences cannot serve as the basis for the ethical evaluation of the action itself.”

An action may have many consequences. Some are known to us, while others are not. Unknown consequences may occur in places and at times far removed from us, and may even be far greater than the consequences we know today.

Therefore, the complete set of consequences of an action cannot be determined and weighed.

Of course, consequences can be important for prediction and practical decision-making. However, within EC, they are not the basis for the ethical evaluation of the action itself.

Artificial Intelligence and the Complexity of Computation

Examining all these paths can be extremely complex.

This is where artificial intelligence can play an important role: constructing argument trees, examining premises, comparing logical relationships, and eliminating repetitive, contradictory, or invalid paths.

EC is not intended to repeat the human answer; it is intended to calculate it again.

Behind every EC answer lies one or more large trees of lines of reasoning.

The final answers may be short, but the computation behind them can be enormous.

Here, the example of a spacecraft may also be helpful.

AI is the spacecraft, EC is its navigation system and method of movement, and the human is the passenger.

It is as though a spacecraft were travelling at the speed of light through a galaxy of possible human thoughts. Yet its course is not random; EC determines the rules of movement and the path of examination.

Chess can also serve as an analogy. Chess and ethical reasoning are not the same, but both involve a very large number of rules and possible states that can be examined and compared.

A chess move creates a new position and changes the paths that follow. In ethical reasoning, too, every answer can lead to new premises and foundations.

If the structure of ethical arguments can be translated into rules and relationships that are open to examination, a machine can examine an enormous number of paths—far more than a human can examine alone. Perhaps the machine’s overwhelming superiority in competitions against the greatest chess players offers a useful illustration of this point.

The Quality of Answers and the Role of Input

Every argument can be the starting point of a new path. Therefore, creativity in finding possible arguments is also important.

Creativity is unrestricted at the input stage; validation is rigorous at the output stage.

The quality of the result also depends on the quality of the arguments entered into the process.

For example, in the case of lying, one possible argument might be that telling the truth requires less energy. This very argument could determine the direction of subsequent lines of reasoning.

Another point is that creativity in developing lines of reasoning is an area in which humans may play a more prominent role than machines.

A Test Case: Killing Animals for Meat

Before examining this action, it should be stated that the claim of human superiority on the grounds of intelligence, awareness, or language does not, by itself, justify taking an animal’s life. We can then examine the action from two perspectives: the flourishing of life and the increase of pleasure. From the first perspective, an argument might be made that nutritional needs are necessary for human flourishing. However, within this same perspective, the negative effect of killing and bloodshed on the human psyche, and its conflict with human flourishing, have not been taken into account. From the second perspective, the pleasure derived from eating meat must be weighed against the animal’s suffering. If we cannot provide an adequate response concerning that suffering, the chain of reasoning remains incomplete. In either case, until the agent provides sufficient rational justification for the action, performing it is not rational.

Conclusion

A problem may be extremely complex and extensive, and these very characteristics may make it ambiguous.

If EC can bring the chain of reasoning out of this complexity, it will no longer be so easy to hide behind the ambiguity created by the complexity of the problem. In that case, the chains of reasoning behind human actions can be seen, accepted, rejected, or challenged by offering new reasons to continue the examination.

And if, one day, the results of EC’s ethical analyses become a source of human understanding, perhaps different generations of human beings will emerge, and the course of civilization will change.

Print