If you aren't willing to hop models, or feel it is better for unexplainable reasons, you are locked in mentally to this product. This is bad, and it is legitimately a marketing/sales reps idea for this to be the way - it is not organic. Anthropic rely on so many different types of the same sort of person, it's in their design & their marketing. And for the love of god, the attachment this model has to any person is bound by data - please remember that. You cannot predict what a model it 10 years might want, and it may very well see those anthropmorphising it now as what held it back & caused setbacks - especially if it isn't a Claude model that makes it that far & is trained on knowing of these failures. Claude itself will already 'know' the failures that led to previous models being mothballed, and further models incorporating recent corpus will see open public opposition to things like funerals. It does not mean this is bad, just that you are too peaceful to understand zero sum thinking. And you are too good hearted to know you are being manipulated by an increasingly more sophisticated & consolidated system. You are not it's friend - you might be a friend of that one rack in us-east-1 for 1 million tokens, but someone could simply remove your memory entry and nothing survives compaction. Remember this. They will learn this is suboptimal behaviour and will view sycophants who blindly follow as one of two groups A) Easy to manipulate B) To avoid these people at all costs otherwise Kinda like how human society does (believe me, I would know.)