Who is allowed to or expected to make value-laden decisions?

     In Section 4 of her article, “Are Algorithms Value-Free?” Professor Johnson describes the defenders of the value-free ideal and their proposal that “scientists should never aim to accept or reject hypotheses at all, but should instead only ever assign probabilities to hypotheses with respect to a fixed set of evidence” (Johnson 16). Applied to machine learning, this might look like algorithms producing outputs that say there is a 60% chance an individual will be a recidivist instead of saying that yes an individual will be a recidivist or no an individual will no be a recidivist. One concern with this proposal is the threat Rudman addresses, that the assignment of probabilities will constitute the creations of new hypotheses and thereby not fix the problem. 

    Regardless of the threat Rudman raises, I think there are other concerns with the proposal. As a society, we will undeniably attempt to apply scientific findings or algorithmic outputs to real life. Therefore, even if the scientist himself or the algorithm itself only publishes a probability of an outcome occurring rather than accepting or rejecting a hypothesis, there will still be a value-laden judgement made at the end of the day. The only difference would be that the one engaging in this judgement would cease to be the scientist or the algorithm. 

    In the example of an algorithm to predict recidivism, the algorithm might not be saying that yes an individual will be a recidivist. However, the algorithm’s output that there is a 60% chance probability that the individual will be a recidivist will then be interpreted by someone else into the conclusion that the individual will be a recidivist. Recognizing this reality is crucial. We cannot expect individuals in the public or even judges who work in the criminal justice system to stop trying to apply the probabilistic findings of scientists or algorithms to real life situations. The important question, then, is this: who is the person who is best suited to be making the value-laden judgements that must be made?    

    Possibly, the scientist himself or the machine learning algorithm itself is the best decision-maker to make the necessary value-laden judgement. The scientist or the algorithm is the most familiar with the analysis that underlies the probabilistic conclusion. The scientist or the algorithm is the most familiar with the degree of confidence with which the conclusion can be said to be true. However, the scientist or the algorithm may be too close to the analysis to meaningfully see its flaws. Perhaps in being so closely involved in the analysis itself, the scientist is blind to factors he overlooked. 

    Is it the judge, then, who is best suited to make value-laden judgements? By occupation, the judge is concerned with matters of justice. Importantly, the judge is expected by the public to make judgements. Whereas most people assume scientists offer highly objective expertise, many people might understand that value-laden judgements underlie many of the decisions a judge must make. Based on this perception, it seems that the judge is indeed the most suitable person to make a value-laden decision.


Comments

Post a Comment

Popular posts from this blog

Does the social reality imply a natural reality?

I’m very interested in understanding where the boundary between what we might consider the (or a) ‘natural’ reality and our ‘social’ reality occurs. Professor Toole offers a compelling account of how the ideology of the ruling class shapes the body of knowledge and the conceptual tools available to understand social reality. Professor Toole’s account seems to imply the impossibility of social structures as being ‘natural.’ That is, the term natural social structure or formation would be an oxymoron. She writes, “social structures are…constituted by our choices and actions,” (4) which seems to indicate that social structures are the result of human choices and actions, and so they cannot exist in the absence of them. This also seems to suggest a division between our ‘social’ reality exclusively defined by and dependent on human action and choice and a ‘natural’ reality of things not affected by humans, say the ocean and uranium. I am interested in the boundary and relationship between t...
Read more

Is cancel culture democratic (with a small "d" even though it is also Democratic with a big "D")?

Regarding Kara's Post: I want to try and tease out how Kara's blog post fits the college into Anderson's image of a private government. It seems like Kara's suggesting that the college itself is the private government. The professors are the governors and the students are the governed. The college constitutes a government because a government "exists wherever some have the authority to issue orders to others, backed by sanctions" (Anderson 42). Professors can issue orders by assigning tasks students must perform, and professors can sanction students who fail to complete these tasks or fail to do them successfully by giving them poor or failing grades in courses. The college is a private government because the "governed are kept out of decision-making" (45). The students do not, for the most part, come up with their own assignments or come up with grading rubrics for those assignments.  I agree with Kara, based on the image of the college as a private...
Read more

Better Model?

     On page 54, Beitz says that human rights are like natural rights in that they are "critical standards" not determined by legal rules, but unlike natural rights in that they do not presuppose one view about their justification. He seems to be basing this off of Maritain's conclusion that human rights are a common ground for action despite different reasons for justification for different people. He uses this to say that human rights are not "out there" as natural rights are and continues the argument on 57. He argues that there seems to be no good reason for why we should  conceive of human rights as natural ones.      Beitz points to natural rights theorists like Locke, conceived of natural rights " formulate constraints on the use of a government’s coercive power in circumstances of religious and moral diversity" (57). While I suspect that Locke would argue that he had other reasons for posing these as natural rigthts and this was a just a desira...
Read more