←back to thread

559 points Gricha | 3 comments | | HN request time: 0.608s | source
Show context
maddmann ◴[] No.46232972[source]
lol 5000 tests. Agentic code tools have a significant bias to add versus remove/condense. This leads to a lot of bloat and orphaned code. Definitely something that still needs to be solved for by agentic tools.
replies(2): >>46233443 #>>46233663 #
1. oofbey ◴[] No.46233443[source]
Oh I’ve had agents remove tests plenty of times. Or cripple the tests so they pass but are useless - more common and harder to prompt against.
replies(1): >>46235744 #
2. maddmann ◴[] No.46235744[source]
Ah true, that also can happen — in aggregate I think models will tend to expand codebases versus contract. Though, this is anecdotal and probably is something ai labs and coding agent companies are looking at now.
replies(1): >>46237595 #
3. oofbey ◴[] No.46237595[source]
It’s the same bias for action which makes them code up a change when you genuinely are just asking a question about something. They really want to write code.