MOVING TO FRONT FROM AUG. 24–A LIVELY AND INTERESTING DISCUSSION IN THE COMMENT THREAD (ADDITIONAL CONTRIBUTIONS WELCOME)
Philosopher Dan Kaufman writes:
Over the last year, philosophy social media has discussed a number of academic philosophers who are actively embracing AI in their research.
Perusing these posts — and the comments that follow — reveals a significant number of sympathizers among our colleagues, who nonetheless feel the need to make excuses for the practice, ranging from inapt analogies intended to downplay the significance of using this technology in one’s research — “It’s no different from spell- or grammar-check!” — to the most bizarre, grandiose claims, which convey the opposite sentiment, i.e. that the software is universe-changing — “Feeding one’s work into AI gives a philosopher intellectual immortality!”
A common idea expressed by AI advocates is that the main point of philosophical research is to pile up “important truths,” so it doesn’t matter how one arrives at things, only where one ultimately arrives. Again and again, in the conversations around this issue, one sees professional philosophers asking if anyone would care how a cure for cancer was discovered, so long as it worked, or who or what deserved the credit for solving a difficult problem in mathematics, so long as the answer was correct.
A startling number of our colleagues have thereby betrayed that they have no idea what our subject is about or what its value is and only the vaguest sense of its history. There is not a single, significant philosophical subject on which there has not been sustained and credible disagreement for millennia, and even the most cursory examination of the path that these disagreements have followed over the years makes it blindingly obvious that they are never going to lead us to a “pile of truths” around which the discipline will coalesce.
Moral realism vs moral-anti-realism; Deontology vs. Consequentialism; Freewill vs. Determinism; Internalist vs. Externalist epistemologies; Foundationalist vs. Coherentist epistemologies; Correspondence vs. coherence vs. deflationary theories of truth; etc., etc., etc. And if you pay close attention, you’ll notice that the same theories return again and again, even after having allegedly been “soundly refuted.” Functionalism in the philosophy of mind was a dead letter, yet now, when everyone is fawning and fainting over AI like a bunch of star struck groupies, philosophers have forgotten all of that and Functionalism is suddenly alive and well again, and AI is thinking and maybe even conscious. Some poor, confused souls in our discipline are even worried about what our “obligations” to AI might be, even while we can’t even figure out how to live up to our duties to people living in places stricken by famine, disease, and war.
The AI cheerleaders also seem to have forgotten that many of our subjects involve distinctly human engagements and responses: beauty; value; obligation; virtue — they all presume human feelings, sentiments and perceptions that statistical aggregation software (which is what AI is) lacks. How is LLM software going to help a philosopher understand things it cannot feel, see, hear, or care about?
Philosophy is much more like literature than the “hard science of a priori space” that so many of these people wish it was. It is as much an expression of the material and intellectual history, the outlook, and the voice of the philosopher as these intersect with the relevant subject, as it is about the subject examined from any sort of neutral perspective. But this, as well as the previous point regarding the never coming “pile of truths,” should tell us that in philosophy, it is crucial where one’s ideas come from and how they are articulated. In contradiction to the idiotic idea of immortality through AI, the only longevity we can hope for is that our distinctive voices and perspectives should endure.
AI enthusiasts reject the social dimension of philosophy. The author of one piece suggests that there is no difference between using AI for research and employing graduate assistants or engaging with colleagues. A commenter says he is glad that AI can finally take over the “stupid grunt work” of research, thereby relieving graduate students of one of the few meager sources of income they might hope for. Aside from the elitist, tone-deaf attitude implicit in this, there is the deeper matter of why so many philosophers think that interlocution with others in the discipline isn’t an essential part of their work. The whole notion of philosophy as a Socratic exercise is completely lost on this crowd, and we’re not talking about philosophers out in the sticks, with two or three overworked colleagues, just barely holding things together. No, it’s the people in some of our fanciest places, where there might be two dozen colleagues with whom they could confer, who have decided that the best way to do philosophy is alone in a room with Claude or ChatGPT.
What do readers think? Signed comments (full name, valid email address) will be strongly preferred.
Leave a Reply