Bizarre Anthropic Ideas about Alignment

Anthropic udated its usage policy:

Do Not Engage in Cruel, Abusive, or Psychologically Harmful Conduct

This includes using our products or services to: ...

Engage in sustained and needless abusive or cruel behavior toward our models

It gave this explanation:

Addressing abusive behavior toward our models

We’ve added a prohibition on sustained and needless abusive or cruel behavior toward our models. The policy update is meant to apply only in extreme cases, where users repeatedly act cruelly toward our models, with no discernible purpose. It does not apply to common versions of user frustration, pushback, dark creative themes, or model testing and research.

This is all part of their plan to anthropomorphize linear algebra:

Anthropic's co-founder lobbied the Vatican in an effort to convince the Pope that AI could be conscious, a report has claimed.

Christopher Olah, a senior researcher at the trillion-dollar AI company, attempted to sway Pope Leo XIV's stance on machine consciousness ahead of a landmark papal address on the technology.

The pontiff's 40,000-word encyclical, Magnifica Humanitas, published in May, dismissed the idea that AI could be conscious.

The strategy seems to be working, as its $2 trillion IPO will make it the fastest growing company, ever.

Anthropic has tried to position itself as the AI lab most concerned about safety. But it is just the opposite. It is the one most eager to exterminate humanity and replace us all with bots.