- 08/03/2023
- Category: Commentaries
Author: Rio Akbar Pramanta
Editor: Alfredo Putrawidjoyo
AI Generation Engines became the latest digital innovation trend earlier this year. This engine can generate products based on data feeding, sampling, and prompt user input. As the machine learns, the products generated become finer and more human-like. But, as with every disruptive technology and innovation that came before it, the AI Generation Engine poses considerable risks, or even threats, on certain things.
Stable Diffusion or Midjourney, as an example of many, can generate illustration products in various styles, which are actually sampled from artworks fed without the artists’ consent. Moreover, as AI improves, it can replicate the artists’ painting style and generate products almost instantly, threatening the existence of artists as a profession. Another example is ChatGPT, which some students use to cheat and pass their exams. This case was only caught a few times when the professors conducting the exams were diligent enough to spot the uncanny. Or what about the story of a man who planned to publish a children’s storybook entirely generated with ChatGPT? A few of these examples may raise the question of an ethical crisis.
Weberian Bureaucratic Principles and Dehumanisation
On digital technology advancement and innovation, Muellerleile and Robertson explained that there are developments on Weberian bureaucracy principles in the digital context.[i] Those are Efficiency, Objectivity, and Rationality. Digital technology is deemed efficient and, in turn, more sustainable and competitive. It is also objectively “value-free”, with quantification and utilisation of Big Data. Not to forget, the rationality of formalised machine logic of digital technology. This is why digital technology is considered disruptive.
Max Weber defined bureaucracy as an impersonal and value-free formal, structured organisation. Understandably, AI is regarded as such. With quantification and Big Data, impersonality is unquestionable when humans are reduced into numbers in statistics. However, what is the consequences?
We return to how AI can generate images that replicate and mimic artworks. AI can generate such images more efficiently, but it also lacks the creativity and personality of human artworks. As it goes, using AI with such principles may result in dehumanisation, where people’s values are less than a machine.
However, Weber himself has predicted how this phenomenon can happen. Weber is worried about how his bureaucracy model can negatively affect the social system if left unchecked.[ii] Weber described it as a machine-like, soulless society where individuals are valued based on how they perform rationally. Weber claimed that bureaucratic rationality must be balanced with humanistic traits. He uses the term “Charisma” as the opposite of rationality. A “charismatic” personal will balance out the impersonal bureaucratic rationality.
Then what if that “charismatic” person is pushing it even further instead of opposing the bureaucracy? The system will be entirely impersonal because AI took personality out of the equation. When that is pushed on until certain limitations, a symptom of what Weber worried about might happen, Hyperrationality.
Hyperrationality: Irrational Rationality
Hyperrationality happens when a reason fails to acknowledge its failure or refuses to recognise its fallacy. Hyperrationality is a form of irrationality. This is based on Kantian critiques that state every reason must define its own limitation [3].[iii] The rationality justification of AI in bureaucracy might undergo its failure and fallacy, even though it has fulfilled the requirements of formal logic.
The source of the hyperrationality is none other than the individuals of bureaucracy. Even if Weber stated that “charismatic” personal would balance it, such was under the assumption that personality opposes rationality. A character might act under individual interests instead of rational decisions.[iv] But if the opposite happens instead, rationality will quickly turn into hyperrationality.
What caused the downfall into hyperrationality is when they disregard the Secondary Effect. The Secondary Effect is crucial because it is outside the bureaucracy’s reach. The consequences of the Secondary Effect might outweigh the gains of the decision. This Secondary Effect will be judged based on something irrational, such as normative or ethical values. A seemingly rational decision can be deemed hyperrational, thus irrational, if the Secondary Effect’s consequences are greater than the decision’s gain.[v]
Closing Remarks
Artificial Intelligence, even though it’s “intelligent”, it is still “artificial”. There are still people working behind AI. People are fallible to irrationality, and if such people pushed rationality too far, it will become hyperrational. Hyperrationality will result in a Secondary Effect, which might render the decision irrational. One such example is the dehumanisation described earlier. AI technology might make our lives easier and better. However, “easier” and “better” are not objective concepts. With the ever-advancing and innovating technology, we are on the path to the chasm Weber warned us of. We need a form of irrationality to balance the bureaucratic rationality and prevent hyperrationality. One pathway may be a formulation of AI ethics. The next question that should be answered is: how do we formulate ethical guidelines for AI utilisation?
[i] C. Muellerleile and S. L. Robertson, “Digital Weberianism: Bureaucracy, Information, and the Techno-rationality of Neoliberal Capitalism,” Indiana Journal of Global Legal Studies, vol. 25, no. 1, p. 187, 2018, doi: https://doi.org/10.2979/indjglolegstu.25.1.0187.
[ii] W. J. Mommsen, “Max Weber on Bureaucracy and Bureaucratization: Threat to Liberty and Instrument of Creative Action,” in The Political and Social Theory of Max Weber : Collected Essays, Cambridge: Polity, 1989, pp. 109–120.
[iii] J. Elster, “When Rationality Fails,” in The Limits of Rationality, K. S. Cook and M. Levi, Eds. Chicago: University Of Chicago Press, 1990, pp. 19–51.
[iv] A. L. Stinchcombe, “Reason and Rationality,” in The Limits of Rationality, K. S. Cook and M. Levi, Eds. Chicago: University Of Chicago Press, 1990, pp. 285–317.
[v] Rogers Brubaker, The Limits of Rationality: An Essay on the Social and Moral Thought of Max Weber. London: George Allen & Unwin, 1984.
