Some people think that LLMs should not be used in technical fields like programming or mathematics, because those fields are actually art. True or not, this kind of argument is not going to be persuasive to the general population. The argument goes something1 like this:
It doesn’t matter how effective LLMs are at technical fields like programming or mathematics, because one of the essential values of those fields is that they’re art. While it looks to laypeople like these fields exist to solve practical problems, what they’re actually for is to produce works of beauty, or to express something distinctively human, or to spend time contemplating the infinite. Simply solving the problems misses the point in the same way that building an infinite poem generator would miss the point of poetry.
I agree that programming or mathematics can be beautiful. I remember being mildly awestruck at learning epsilon-delta proofs, and I have written a lot about the kind of software I think is elegant. Since there’s no clear definition of art or of aesthetics in general, it’s plausible to me that almost everything could have an artistic dimension to it.
All of this is compelling to software engineers, many of whom (like me) find some programs beautiful and are flattered to think of themselves as artists. But I suspect it’s spectacularly uncompelling to onlookers, because it undermines the implicit reasons that they have for supporting technical fields. Many people who are not programmers or mathematicians think of those fields like this:
- I enjoy the fruits of technology: modern medicine, transport, mobile phones, and so on
- While I don’t understand the exact relationship, I know that technical fields like mathematics and programming are essential to sustaining and advancing these technologies
- Therefore I’m happy to respect and support mathematicians in their mysterious research work
- And I’m broadly okay with2 software engineers drawing healthy salaries for the same reason
In other words, regular people3 think mathematicians (and to a lesser extent programmers) are valuable because they think science is valuable, and they think science is valuable because they think technology is valuable. They are willing to accept that technical fields can be art only insofar as that eventually helps researchers have the kind of creative insights that advance science: as Rikab sarcastically puts it, that advance the knowledge frontier.
If you go to those non-technical people and you say “LLMs can accelerate problem-solving in these fields, but I’m strongly against using them because the real point is to produce great art”, they will be very unhappy. They might reasonably feel cheated, in the same way as if they hired someone to renovate their kitchen who insisted in doing it all with hand tools in order to experience “the joy of creation”. This argument — we should stop using LLMs in technical fields because they’re art — is going to be even less compelling to technical employers.
Engineers and mathematicians who want to defend the artistic value of their field will need more compelling arguments4 than just appealing to the value of art. We will need to find some way to connect the value of art with the practical technological value that motivates public interest in our fields. For instance, we could argue that:
- Great software products must rely on elegant abstractions, which need a human creative touch: simply piling on working code will hit a wall past a certain level of complexity
- LLMs can solve problems, but actual progress in mathematics depends on creating new abstractions, which is a more creative task that LLMs are not yet good at
- Progress in technical fields is uneven and relies upon inspiration, which needs creative hammock time and distinctively human understanding
- LLMs can produce working software, but truly great software requires human inspiration5, and great software is what advances the field
I don’t know if I agree with all of these points. I haven’t yet seen an LLM-created codebase truly hit a wall (at least, not this year), and I don’t see a reason why LLMs should be incapable of creating new abstractions or having human-like inspirations. I also don’t think great products have to have great code: why wouldn’t workmanlike code in service of a great idea be enough? Does Fallingwater have to have beautifully installed6 insulation?
Still, any argument we make to our employers or the general public about art is going to have to try and make this case. Instead of saying that human-made technical work is valuable because it’s art, we need to make an affirmative case for why art can make technical work more effective. Otherwise we’re conceding the main reason why the general population thinks our work matters.
-
As usual when I’m trying to avoid strawmanning a point of view I don’t hold, please email me if you agree with the quoted paragraph and consider it a fair representation of your position. If nobody writes in to tell me they agree, I will reword the paragraph.
↩ -
Leaving out some specific concerns about the health of social media, short-form video, gambling apps, and so on.
↩ -
Is there actual evidence for this claim I’m making, other than it just seeming kind of obvious? I couldn’t find anything definitive, but the UK CaSE study (see “The balance of fundamental and applied research”) and this paper are suggestive.
↩ -
This is not a criticism of the Glyph post, which is aimed at other engineers, and does include an argument that artistically-produced software is better for users.
↩ -
I take this to be the argument in the Glyph post.
↩ -
Ironically, the house does actually have serious mildew problems. The poor construction quality does not seem to have affected its status as art.
↩