When does "no" mean no? Insights from sex robots

> Part of what makes sex robots ideal in the context of the current work is that people's attribution of moral standing to robots is highly manipulable, as people's moral intuitions regarding robots are much weaker and more flexible than those regarding humans ... We can therefore leverage this flexibility of people's intuitions about AI robots in cases of sexual assault.

Sounds ideal from the perspective of getting the result you are looking for. I'd take these attributes as signs that the intuitions are weak and arbitrary.

The justification seems to be:

> these disparities led to the emergence of an alternative view which suggests that social robots are interpreted merely as composite depictions of social agents, rather than entities that lie on a continuum of mind perception ... If people view female-like robots as depictions of women, for example..

That's a big if, would seem to me there should be a lot more evidence for this (and any nuances of the process) before basing further studies on it?

Now that you have a "highly manipulable" model, and a theory that these result apply to real women, you can manufacture any moral perspective on women you choose.

This is studying under what conditions sexual assault may seem morally acceptable depending on the 'victim.' To my mind, this is perverse; sexual assault should be morally unacceptable regardless of who (or what) the victim is. The immorality lies not only in the violation of the victim, but also in the lack of control of the immoral impulse in the perpetrator. This study seems to implicitly accept that sexual assault can be more or less moral depending on the circumstances.

I find the rambling writing style hard to understand. Concise and direct would be a big improvement to the paper.