Did Anthropic not recently hire some high-ranking Vatican officials to directly influence the training and ethical rules of their model? Presume they talk deeply about religion to train the model ethical standards. Should have take the offer from Anthropic to influence the training imo.

The author states that he did not join because he did not believe he would have influenced them.

I feel from reading this that Anthropic would benefit from the work of the Nobel Laureate Herb Simon and his book the Philosophy of the Artificial.

So Simon argues that whilst traditional Meta Physics is about physical properties, with artificial creations it is different, what they do is more important than how they are made.

It more important that a watch tells the time than that it contains cogs and crystals.

A computer is defined not by the silicon that it made of but by the software it runs.

So with AI what is important is its function. We don’t need understand how the neurons are wired together