>>514801821
You can't know what the AI has as a relevant probability; you couldn't without being able to peer inside it, and the fact it isn't giving you the answers you want means that no, in fact, your concept of what is a probable answer is not what the actual probable answer is for the AI.

>even though I included the manufacturer in my original search parameter.
That often doesn't matter to the AI. Everything you say to it is not part of its training data. It doesn't learn from conversations with you. As soon as you create a new instance it will be as if your previous inquiries never existed.

It's training data, the topology of its 'brain', is set in stone. There is a very limited amount of wiggle room you can get out of specific prompting, but the fundamental shape of its mind is immutable without a new training run, which means it can be wildly and completely incorrect and tell you total nonsense that isn't even remotely related to what you're asking about.

Furthermore, the google search results AI in particular is also a bottom of the barrel model, probably the lowest quality you could possibly interact with, because it's meant to be cheap and fast to run, not accurate (and expensive).

Pic related for example, is how the free version of grok answers your question.