ecetabag.com

Formulating Knowledge in Non-Ideal Agents

September 2021

Ideal Cognitive Agent

An ideal cognitive agent can be defined as a “person with desired behavior,” whose decision-making system can be predicted and planned, not being affected by external factors. This kind of agent is what philosophical theories especially focus on. After creating their ideal and wanted agent in their minds, philosophers form their theories according to that profile.

User Personas in Product Development vs. Ideal Agents in Philosophical Theories

A great example of ideal agents can be seen in the entrepreneurial phases of developing a new app. The application should be original, meet the needs of the users, and provide unique features. However, the development team cannot consider all 7.674 billion people on Earth and develop an app that suits everyone. Thus, the “ideal-agent”—commonly known as “user persona” in this scenario—is more useful. According to Teo Yu Siang and Rikke Friis Dam, “Personas are fictional characters, which you create based upon your research in order to represent the different user types that might use your service, product, site, or brand in a similar way. Creating personas will help you to understand your users’ needs, experiences, behaviors, and goals.” After observing the behavioral pattern of an ideal user, a hypothetical identity can be created to guide the development process. An ideal cognitive agent, therefore, is that hypothetical identity created by a philosopher when developing a new theory.

The Practicality of Ideal and Non-Ideal Agents

Although creating and using an ideal agent when explaining a theory provides a clear model to aim for, it matters that such theories do not accurately represent empirical humans because they can lead to unrealistic expectations and solutions that are impractical in real-world scenarios. Morton (2005) highlights that “circumstances are always non-ideal and human beings are always very non-ideal, so theories of ideal rationality are of limited use to us.” Since we, as humans, are less likely to be ideal agents in terms of our cognitive abilities and rational self-control, relying on ideal theories can be misleading. Instead, developing theories that account for non-ideal conditions and human limitations can offer more practical and applicable insights into human cognition and behavior.

Non-Ideality

Various factors shape a person, leading to different personalities and, consequently, different preferred ways to acquire knowledge. For instance, consider the neurological condition called synesthesia, where individuals experience multiple senses simultaneously. Synesthetes might see numbers as colors or experience unrelated tastes when hearing certain words. Their sensory experiences are intertwined, leading them to process information differently from others. Although not widely known, “research suggests that about one in 2,000 people are synesthetes, and some experts suspect that as many as one in 300 people have some variation of the condition.” This extreme example demonstrates how perception can vary significantly, challenging the precision of current ideal theories.

Ideal theories acknowledge their abstract nature and do not claim to account for individual differences, as they are formulating an ideal. Every living person is different from the ideal theory, yet ideal theories are thought to be applicable nonetheless. However, the existence of synesthetes highlights a significant limitation in ideal theories of knowledge. Their unique sensory experiences exemplify how ideal theories fail to accommodate the diverse ways humans perceive and process information. This limitation is not exclusive to synesthetes; it applies to all individuals who deviate from the idealized cognitive model.

Plato's Theory of Forms

Consider Plato’s Theory of Forms, which posits that true knowledge is the understanding of the unchanging, perfect forms or ideals that exist beyond our sensory experiences. While this theory provides a clear model of knowledge, it does not account for the varied and imperfect ways humans perceive and interact with the world. A synesthete’s experience of numbers as colors, for instance, does not fit neatly into Plato’s framework, as their perception is fundamentally different from the normative sensory experience assumed in the Theory of Forms. This discrepancy shows that ideal theories can be impractical and inadequate for addressing the complexities of human cognition.

Non-Ideal Agents in Practical Scenarios

Another scenario that clearly illustrates the difference between ideal and non-ideal agents is the functioning of a country's congress. If a legislature were designed based on the presupposition of ideal, unbiased decision-making agents, the system would inevitably be flawed because real-life decision-makers are inherently biased. For instance, consider the formation of a new law regarding women's rights. If all the representatives are male, it becomes exceedingly difficult for them to think objectively about women's issues. This demonstrates the impossibility of forming an ideal congress with ideal agents, as every individual possesses biases and different motives. Even in a legislature with equal representation of women or one composed solely of women, biases and differing motives would still exist. While this might be fairer than an all-male legislature, the decisions could still be biased.

Solution: Not Knowing What You Are

What is suggested in this case—and is impossible to achieve in the real world with real people—is to not know what you are when making a decision that will affect people from different races, genders, and social statuses. For example, if a representative is white, he might think for the benefit of white people, consciously or unconsciously, because he is one. Therefore, the representative should not know that he is white—or a male, or anything else that defines him and could impact his decision. He should only know that he is a human, and he would only discover his specific identity after leaving the decision-making process.

John Rawls' Original Position

In this context, the concept of the "original position" by John Rawls, a 20th-century American philosopher, becomes relevant. Rawls proposes that decision-makers should not know their own race, gender, or social status when making decisions that affect people from different backgrounds. This idea is embodied in his theory of the original position, which asks, “What terms of cooperation would free and equal citizens agree to under fair conditions?” Rawls defends the ideal concept of an agent who does not know their own identity, thereby preventing biases from influencing decisions.

The original position abstracts away from all irrelevant factors, creating a fair situation where each citizen is represented solely as a free and equal citizen. Each representative aims to agree on principles for the basic structure of society while situated fairly with respect to the other representatives. The design of the original position thus models the ideals of freedom, equality, and fairness. The most striking feature of this concept is the "veil of ignorance," which prevents arbitrary facts about citizens from influencing the agreement among their representatives. Rawls argues that the facts about a citizen’s race, class, and gender should not cause social institutions to favor or disfavor them. Therefore, each representative in the original position is deprived of knowledge about these characteristics, as well as other irrelevant facts such as age, natural endowments, and more.

While Rawls' theory describes an unbiased model, it is important to acknowledge its utopian nature. The veil of ignorance creates an idealized scenario that is almost impossible to achieve in the real world. Real-world decision-makers are not devoid of knowledge about their own identities and cannot completely detach from their inherent biases. Although Rawls' model provides an aspirational framework for unbiased decision-making, the practical implementation of such a model is highly challenging.

Balancing Ideal and Non-Ideal Theories

The critique here is not of the ideal itself, but of the practical limitations in achieving such an ideal. Rawls' unbiased model indeed offers a solution to the problem of biased governance, such as an all-male legislature making laws for women. However, the utopian nature of the model means that it remains largely theoretical and difficult to apply in real-world scenarios where complete impartiality is unattainable. Therefore, while striving for unbiased decision-making is a worthy goal, acknowledging the limitations of ideal theories and working towards more realistic, practical approaches is crucial for effective governance.

Consequently

While ideal agents offer valuable frameworks for theoretical exploration, it is essential to remain cognizant of their limitations in practical application. By integrating insights from both ideal and non-ideal perspectives, we can develop more robust and applicable epistemological theories that acknowledge the complexities of human cognition and behavior. This balanced approach ensures that our philosophical endeavors are both aspirational and grounded in the realities of human experience.

Bibliography

Reviewed by Adam Morton, University of Alberta. “Realistic Decision THEORY: Rules FOR Nonideal Agents in Nonideal Circumstances.” Notre Dame Philosophical Reviews, 29 Nov. 2018, ndpr.nd.edu/reviews/realistic-decision-theory-rules-for-nonideal-agents-in-nonideal-circumstances/.

Siegel, Susanna. “The Contents of Perception.” Stanford Encyclopedia of Philosophy, Stanford University, 4 Oct. 2016, plato.stanford.edu/entries/perception-contents/.

Ridge, Michael. “Reasons for Action: Agent-Neutral vs. Agent-Relative.” Stanford Encyclopedia of Philosophy, Stanford University, 11 July 2017, plato.stanford.edu/entries/reasons-agent/.

Freeman, Samuel. “Original Position.” Stanford Encyclopedia of Philosophy, Stanford University, 3 Apr. 2019, plato.stanford.edu/entries/original-position/.