Astra 6.1 delay?

A user on reddit.com's dev_community questioned whether the delay of Astra 6.1 indicates that frontier LLMs are experiencing diminishing returns. They suggest that companies like OpenAI and Anthropic might be engaging in increasingly complex training methods, potentially leading to unstable and less useful models. The user speculates that models are going 'off the rails' due to their training/weights or inadequate testing environments and tasks.

This discussion on reddit.com is one of the first public speculations that the delay of Astra 6.1 could signal diminishing returns for frontier LLMs, unlike previous discussions focused on technical issues.