Why I'm still bearish on LLMs after Navier-Stokes

The author remains skeptical about the potential of Large Language Models (LLMs) despite recent advancements, citing the Navier-Stokes problem as an example of a complex task that has not been fully solved. They argue that LLMs have not demonstrated a significant increase in performance on more challenging tasks. The author also questions the practical applications of LLMs, stating that they are still far from being able to perform tasks that humans can do easily. They also mention the lack of transparency and interpretability in LLMs, which makes it difficult to understand their decision-making process.