Debating Whether AI is Conscious Is A Distraction from Real Problems

Link post

Hi LW, this article has put some doubt in my mind as to whether researching AI alignment is worthwhile, or a frivolity that detracts from more pressing issues such as mitigation of harmful biases.

[Discussing AI consciousness] is a distraction, when current AI systems are increasingly pervasive, and pose countless ethical and social justice questions that deserve our urgent attention

I would like to hear some opinions from this community on the sentiment expressed in the above quote.

First, though, an introduction, or why this has prompted me to emerge from my decade-long lurker state:

I am a computational linguistics MS student approaching my final semester and its accompanying research capstone. My academic background is in linguistics, not computer science, philosophy, math, or anything that would actually be useful for doing serious AI research. c; Nevertheless, I want to try to help, as best I can.

For my capstone, I have been considering continuing a previous bias mitigation project (that I have lost most of my enthusiasm for due to group project burnout), or using LLMs to extract moral themes from literature (an idea my thoughts keep coming back to) in the hopes that the ability to encapsulate the human values expressed in a text may someday be useful to the AI alignment task. The thought process behind this was, although human beings can’t even agree on human values, we all somehow manage to come up with a sense of morality based on our cultural upbringing, and one common form of character education for children is storytelling (Aesop’s fables, Narnia...) If human children can learn virtues from fiction, why not AI?

The part of Lemoine’s dialogue with LaMBDA where it succinctly summarizes the moral themes of Les Mis made me think this may be a good idea to play around with. [Note, just to be extremely clear, I in no way imagine LLMs to be sentient- Lemoine is misguided by his religious thinking. But I do think LLMs are likely to be a component of a future system that could become conscious.]

However, criticism such as the above article makes me feel I may just be wasting my time selfishly, and I should go back to doing sociolinguistics rather than frivolous literary analysis, even if Aesops are more fun.

I would love to know your thoughts, whether about my personal predicament or the article in general.