From the title, I was expecting this to be the result of a jailbreak of Claude. I was pleasantly surprised that it is an Anthropic employee disclosing and explaining the rationale for it.
Yup, a very refreshing perspective considering the other major players' seemingly pointless secrecy about it.
Amanda is one the most talented researchers at Anthropic. Truly an honor to work with her.
She stated "Let me break it down." Did she break it down somewhere? How can we see it without X account?
IMO a better link - Thread Reader is more suited for actually reading something like this, even if you're logged in.
My priors were that Anthropic is the most left leaning & most secretive of the big labs. So this is quite surprising!
I just tested Claude and it failed my alignment test to the point it would violate international conventions on genocide before it would help a conservative.