The post asks whether the responses using the latest version of llama_index will be more accurate than those from version 0.6.35. In the comments, a community member suggests that the accuracy is likely to be about the same, as there were only minor changes to the GPT-3.5/4 prompts. The real changes in accuracy, the community member explains, come from how the index/query engine is configured and how OpenAI updates their language models over time.
