4 ms·

I recently witnessed one such potential fuckup. The AI had written functioning code, except one of the business rules was misinterpreted. It would have broken i

I recently witnessed one such potential fuckup. The AI had written functioning code, except one of the business rules was misinterpreted. It would have broken in a few months time and caused a massive outage. I imagine many such time bombs are being deployed in many companies as we speak.

tombert 9mo ago
Yeah; I saw a 29,000 line pull request across seventy files recently. I think that realistically 29,000 lines of new code all at once is beyond what a human could understand within the timeframe typically allotted for a code review. Prior to generative AI I was (correctly) criticized once for making a 2,000 line PR, and I was told to break it up, which I did, but I think thousand-line PRs are going to be the new normal soon enough.
aurareturn 9mo ago
That’s the fault of the human who used the LLM to write the code and didn’t test it properly.
- tombert 9mo ago
  Exhaustive testing is hard, to be fair, especially if you don’t actually understand the code you’re writing. Tools like TLA+ and static analyzers exist precisely for this reason. An example I use to talk about hidden edge cases: Imagine we have this (pseudo)code fn doSomething(num : int) { if num % 2 == 0 { return Math.sqrt(num) } else { return Math.pow(num, 2) } } Someone might see this function, and unit test it based on the if statement like: assert(doSomething(4) == 2) assert(doSomething(3) == 9) These tests pass, it’s merged. Except there’s a bug in this; what if you pass in a negative even number? Depending on the language, you will either get an exception or maybe a complex answer (which not usually something you want). The solution in this particular case would be to add a conditional, or more simply just make the type an unsigned integer. Obviously this is just a dumb example, and most people here could pick this up pretty quick, but my point is that sometimes bugs can hide even when you do (what feels like) thorough testing.

tombert 9mo ago

Yeah; I saw a 29,000 line pull request across seventy files recently. I think that realistically 29,000 lines of new code all at once is beyond what a human could understand within the timeframe typically allotted for a code review. Prior to generative AI I was (correctly) criticized once for making a 2,000 line PR, and I was told to break it up, which I did, but I think thousand-line PRs are going to be the new normal soon enough.

aurareturn 9mo ago

That’s the fault of the human who used the LLM to write the code and didn’t test it properly.

tombert 9mo ago
Exhaustive testing is hard, to be fair, especially if you don’t actually understand the code you’re writing. Tools like TLA+ and static analyzers exist precisely for this reason. An example I use to talk about hidden edge cases: Imagine we have this (pseudo)code fn doSomething(num : int) { if num % 2 == 0 { return Math.sqrt(num) } else { return Math.pow(num, 2) } } Someone might see this function, and unit test it based on the if statement like: assert(doSomething(4) == 2) assert(doSomething(3) == 9) These tests pass, it’s merged. Except there’s a bug in this; what if you pass in a negative even number? Depending on the language, you will either get an exception or maybe a complex answer (which not usually something you want). The solution in this particular case would be to add a conditional, or more simply just make the type an unsigned integer. Obviously this is just a dumb example, and most people here could pick this up pretty quick, but my point is that sometimes bugs can hide even when you do (what feels like) thorough testing.