Colin's explanation is partially correct. Google has those results because they have very little training data from reddit that corresponds to those questions. This means a single shitpost may be matched to your answer. What you need is a layer that discerns how legitimate the question is before getting any answer. Google is behind. OpenAI has that layer and probably multiple smaller models it can kick you too depending on that judgement.
@furtsmagee15134 ай бұрын
#justiceforscott
@furtsmagee15134 ай бұрын
If chevron wins arbitration BDE is legally required to buy HESS toys trucks for the show