@jornfranke@social.anoxinon.de
Post #2916526
2026-05-14 19:20 UTC
@shadow53@floss.social I would ask to measure that it indeed brings improvement and have additional validation (unit testing, static code analysis, fuzzing etc.) that it does not decrease quality. Static code analyzers give you also statistics on duplicated code and code complexity. They can be indicator that things go wrong. Additionally, you can collect number of incidents (if LLM would work they should go down).
Replies (1)
-
@jornfranke@social.anoxinon.de 2026-05-14 19:26
@shadow53@floss.social Then, there is the cost risk - all of them make a big loss at the moment (https://www.wheresyoured.at/premium-ais-circular-psychosis/). This means you will in the very near future pay 100-1000x more money. On top of that you will spend even more money because the more code you generate the more token you have to spend in the future for AI to go through your code. The cost risk is real: https://finance.yahoo.com/sectors/technology/articles/ubers-anthropic-ai-push-hits-223109852.html