id: 740
uid: hrP3o
insdate: 2026-03-12 13:05:10
title: Are LLMs not getting better?
additional: 🚀 Large Language Models (LLMs) are stagnating in terms of performance improvement, with recent benchmarks showing little to no progress, which raises concerns about their potential to deliver practical value in real-world applications, such as enhanced language understanding and generation capabilities. This stagnation affects their ability to accurately process and respond to complex queries, ultimately impacting their usefulness in industries like customer service and content creation.
category: Hacker News
md5:
guid: https://news.ycombinator.com/item?id=47349334
source_url: https://entropicthoughts.com/no-swe-bench-improvement
updated:
image:
author_name: 4diii
author_link: