AI-generated code is meant for the machine, or for the author/prompter. AI-generated text is typically meant for other people. I think that makes a meaningful difference.

▲

ripe 5 hours ago | parent | next [-]

Code can be viewed as design [1]. By this view, generating code using LLMs is a low-effort, low-value activity.

[1] Code as design, essays by Jack Reeves: https://www.developerdotstar.com/mag/articles/reeves_design_...

▲

acedTrex 5 hours ago | parent | prev | next [-]

Compiled code is meant for the machine, Written code is for other humans.

▲

gordonhart 5 hours ago | parent | next [-]

For better or worse, a lot of people seem to disagree with this, and believe that humans reading code is only necessary at the margins, similarly to debugging compiler outputs. Personally I don't believe we're there yet (and may not get there for some time) but this is where comments like GP's come from: human legibility is a secondary or tertiary concern and it's fine to give it up if the code meets its requirements and can be maintained effectively by LLMs.

▲

threetonesun 4 hours ago | parent | next [-]

I rarely see LLMs generate code that is less readable than the rest of the codebase it's been created for. I've seen humans who are short on time or economic incentive produce some truly unreadable code.

Of more concern to me is that when it's unleashed on the ephemera of coding (Jira tickets, bug reports, update logs) it generates so much noise you need another AI to summarize it for you.

	▲	gordonhart 3 hours ago \| parent [-]
		The main coding agent failure modes I've seen: - Proliferation of utils/helpers when there are already ones defined in the codebase. Particularly a problem for larger codebases - Tests with bad mocks and bail-outs due to missing things in the agent's runtime environment ("I see that X isn't available, let me just stub around that...") - Overly defensive off-happy-path handling, returning null or the semantic "empty" response when the correct behavior is to throw an exception that will be properly handled somewhere up the call chain - Locally optimal design choices with very little "thought" given to ownership or separation of concerns All of these can pretty quickly turn into a maintainability problem if you aren't keeping a close eye on things. But broadly I agree that line-per-line frontier LLM code is generally better than what humans write and miles better than what a stressed-out human developer with a short deadline usually produces.

▲

hinkley 5 hours ago | parent | prev [-]

And Sturgeon tells us 90% of people are wrong, so what can you do.

▲

philipp-gayret 5 hours ago | parent | prev [-]

Compiled natural language is meant for the machine, Written natural language is for other humans.

	▲	CivBase 4 hours ago \| parent [-]
		If AI is the key to compiling natural language into machine code like so many claim, then the AI should output machine code directly. But of course it doesn't do that becaude we can't trust it the way we do a traditional compiler. Someone has to validate its output, meaning it most certainly IS meant for humans. Maybe that will change someday, but we're not there yet.

▲

jvanderbot 5 hours ago | parent | prev | next [-]

This is precisely correct IMHO.

Communication is for humans. It's our super power. Delegating it loses all the context, all the trust-building potential from effort signals, and all the back-and-forth discussion in which ideas and bonds are formed.

▲

askvictor 37 minutes ago | parent | prev | next [-]

At the same time, AI-generated code has to be correct and precise, whereas AI-generated text doesn't. There's often no 'correct solution' in AI-generated text.

▲

ginsider_oaks 3 hours ago | parent | prev | next [-]

> Programs must be written for people to read, and only incidentally for machines to execute.

from the preface of SICP.

▲

everforward 4 hours ago | parent | prev [-]

A lot of writing (maybe most) is almost the same. Code is a means of translating a process into semantics a computer understands. Most non-fiction writing is a means of translating information or an idea into semantics that allow other people to understand that information or idea.

I don’t think either is inherently bad because it’s AI, but it can definitely be bad if the AI is less good at encoding those ideas into their respective formats.