More Comments: Smart Search Synopsis

Dave Hooton
Dave Hooton MVP Posts: 35,763

1. A good response and Synopsis, and reproducible.

image

2. This question first elicited an error "try again" response, then the three different Synopses below:

image

image

image

Note the different attributions in the last two Synopses as well as differences in the first and last sentences.

Dave
===

Windows 11 & Android 13

Tagged:

Comments

  • MJ. Smith
    MJ. Smith MVP Posts: 53,396

    bumping in order to get a response

    Orthodox Bishop Alfeyev: "To be a theologian means to have experience of a personal encounter with God through prayer and worship."; Orthodox proverb: "We know where the Church is, we do not know where it is not."

  • Adam Borries (Logos)
    Adam Borries (Logos) Community Manager, Logos Employee Posts: 881

    With generative AI, it's not necessarily expected to get a precisely consistent response. The top search results from which the Synopsis is generated should be consistent, but the response itself may change subtly. This is especially true as we fine-tune the prompt in development. But even apart from development, some differences are expected: even though we prompt the AI to stay as close to the source material as possible, there is always a degree of variability with an LLM. In that respect, the differences in last two Synopses are not surprising. 

    Having said that, it would be surprising to get a Synopsis that is substantially different -- a difference in substance, not merely in wording. I might put your second screenshot in that category, as it doesn't provide any sources, and (in my opinion) doesn't answer the question particularly well. 

    Is that the response you were looking for, Dave, or was there a different question or concern?