Comment by mo_42
1 day ago
Lately, I've been thinking that LLMs will lift programming anyways to another level: the level of specification in natural language and some formal descriptions mixed in. LLMs will take care of transforming this into actual code. So not only users don't care about programming but also the developers. Switching the tech stack might become a matter of minutes.
How will that work out?
If it simply generates code from natural language then I am still fundamentally working with code. Aider as an example is useful for this, but anything that isn't a common function/component/class it falls apart even with flagship models.
If I actually put my "natural language code" under git then it'll lack specificity at compile time likely leading to large inconsistencies between versions. This is horrible user experience - like the random changes Excel makes every few years, but every week instead.
And everyone that has migrated a somewhat large database knows it isn't doable within minutes.
I don't think one would put only the specification in Git. LLMs are not a reliable compiler.
Actual code is still the important part of a business. However, how this code is developed will drastically change (some people actually work already with Cursor etc.). Imagine: If you want a new feature, you update the spec., ask an LLM for the code and some tests, test the code personally and ship it.
I guess no one would hand over the control of committing and deployment to an AI. But for coding yes.