Negative users dismissed Anthropic's anonymization of partner names in the new Claude Memory System as PII paranoia, a preventable technical failure, or possibly fabricated without proof.
Based on 3 visible X reactions from 12 accounts; directional sample.
Ask a question below.
Published answers will appear here.
The new in-house system aims to prevent memory hallucinations.
@xlr8harder its faked, or he has terrible history - claude will save partners names just fine - and i dont see no proof really, this could so easily be faked output
@xlr8harder PII detection is quite deterministic and around before AI was cool. How can they stuff that up like that. SMH.
@PaulGugAI Yeah, I imagine it's PII-related paranoia.
The old summarizer kept "remembering" hallucinated facts
i've never seen a company that trusts its customers less
Negative users dismissed Anthropic's anonymization of partner names in the new Claude Memory System as PII paranoia, a preventable technical failure, or possibly fabricated without proof.
Based on 3 visible X reactions from 12 accounts; directional sample.
Ask a question below.
Published answers will appear here.