blog:2024:1113_thinking_reflective_judgment_and_ai_design [Witapedia]

This page is read only. You can view the source, but not change it. Ask your administrator if you think this is wrong.
====== Thinking, Reflective Judgment and AI Design ======

There is a double aspect to humans, participation, and the design of artificial intelligence systems. It's clear in Hannah's work and the philosophers she draws upon, that "thinking" is only a human activity. Of course, in their time, there may have been limited technology, certainly in Socrates' but also Hannah's time. And so her studies and critiques of technical rationality could only be based on advanced techniques and rules ([[:technical rationality]]) used by people to make decisions. 

===== More =====


However, as Feenberg comments, her theory doesn't take into account the word or concept of technology, even though her mentors and contemporaries did, such as Heidegger. And so as we turn to the artificial intelligence system used by the schools to determine at-risk people who require funding and lunches, Arendt and Hedeigger's critiques are not just about the over-creeping of technical specificity. Taking into account participation as "part taking", that the people most affected by the AI design and the resulting funding decisions are absent from any real conversations about the systems produced to categorize them speaks to a whole range of critical theoretical worries. 

More importantly, that machines can't think is an important insight, especially when Hannah draws upon Socratic ideas about thinking as having **no** particular end, knowledge, or product.  The examined life is worth living only because it makes us human. Perhaps Socrates is out-dated because his philosophy can't take into account the advances in artificial intelligence and its "thinking" role (need to comment on his suspicion of the printed word).
 
However, Hannah's examination of the German officer eichman and his lack of thinking through simply following rules, in his case rules produced by others, perhaps illustrates a number of things, including a problem introduced 
by philosophy, which created a specialized group of individuals to think, while the vast majority are left to take care of the material and social needs of society so that these individuals can do the thinking. Her critique of eichman is that he didn't think, the way everyone should think, in dialogue with that inner voice, which speaks and challenges us to consider the fraughtness of any situation. There is a nice quote from Socrates effectively saying that it was much better to be at peace with that inner voice, even if it meant that the society around us is discordant and chaotic.  {Fair to say that with Trump's election, we live in both} 

She then goes on to lay out very specific criteria for reflective judgments arising from our impressions of particular experiences, that achieve wisdom as per Kant, through collective conversation and discernment across individuals. The only purpose is to achieve an ever-growing enlarged mentality (see [[:Feenberg]]' use of this), holding the many diverse perspectives in our thinking and thinking, at the same time. This is a difficult and arduous task, given our dominant motivation to finish quickly, and strive towards results. 

From reflective judgments, there is also this critique of determinant judgment through pre-existing categories. The critique is not so much that we should never use pre-existing categories, but like the model builder talking with his boss who comments about his model factory: "great model Bingham, but where are we going to find the tiny people?", we often overwhelm and force any details and particularities of what's in front of us into these pre-given categories. We could argue that most of what we see in the educational AI example are determinant judgments from the imposition of categories, measurement, and data onto particular people and places, with very-real consequences without appeal.  You can heard it: "you are excluded because the data and model indicate that you are find and not in-need of funding". 

In the background are questions of why? Why did this model emerge, what are the assumption categories and limitations used to evaluate success and the processes to achieve it?  How would the beneficiaries (victims) experience it, and its effects?

{{tag>design_paper}}