It can be useful, but it can miss some things, or even make stuff up. A test of several questions posed to three models.
I gather from this that its flaws are within the laws of averages(averaging information) and so not very trustworthy for finding insights.
Yes, but.. the better ones are a reasonably good jump start for those who (as I think will be more common) start first w/AI rather than search--and also don't have the ability/effort to assess sources.
I gather from this that its flaws are within the laws of averages(averaging information) and so not very trustworthy for finding insights.
Yes, but.. the better ones are a reasonably good jump start for those who (as I think will be more common) start first w/AI rather than search--and also don't have the ability/effort to assess sources.