In the latest breakthrough from the world of artificial intelligence, researchers from the University of Bath and the Technical University of Darmstadt have provided compelling evidence that large language models (LLMs), despite their linguistic prowess, are far from posing any existential threat to humanity. The findings, unveiled at the 62nd Annual Meeting of the Association for Computational Linguistics (ACL 2024), emphasize that while LLMs continue to improve in generating sophisticated language and following detailed instructions, they do not possess the inherent ability to acquire new, complex reasoning skills independently.