Comment by iceKirin

I feel that the recent iterations of LLM haven't provided an intuitive qualitative leap. Have they entered a bottleneck period so quickly?