Novelty is one of the main concerns when coming up with an idea. Especially on this platform, we invest quite a lot of time testing if our ideas have already been proposed/implemented somewhere. We also invest time to test the ideas of others in that regard.
What if there was an automated system to do this? I imagine it would look like a search engine just focused on that exact purpose. It wouldn't show you related results based on keywords and other factors but instead would analyze all the related extensive data it could get access to and provide a summary for you to see. The summary might include the number of times a similar concept is mentioned on the web, when was the first time something similar was mentioned, announcements of the attempts to implement this in practice, was it actually implemented or just anticipated to be, etc. The system could give a score in each of those categories and then some general score of novelty.
AI would provide links in the summary and you could re-visit some of those places on the web yourself to test the content and see if the software was accurate in determining the similarities between your idea and other concepts. You could also use filters to test just the specific aspects of novelty, e.g. practical realization or the theoretical concept itself, etc.
This would be useful for many people in many fields. If such a system were trustworthy it could be used as an objective measure by people writing scientific articles and for similar purposes where a rational, objective evaluation of novelty is important.