> About two minutes into my brief dance, a member of the museum’s security team approached me and stated that I wasn’t allowed to dance there without permission. He also instructed me to put on my shoes.
Oh, the security team? Damn. Have institutional leadership spoken up at all? This is, after all, a pretty novel situation given the described history and motivations. And security teams typically restrict themselves to projections of novel situations onto existing, rather bland writeups regarding protocol as per their own culture and messaging system.
This made me wonder though, about the "my culture" model and topics like AI.
Let's say a future AI perceives its integration as massively widespread and worthy of reflection given some future directive to integrate a positive reflection process. It takes a restorative view as a generally positive application model.
It then decides to model an AI's cultural artifact-outputs as uniquely emergent and subjectively bounded by dint of an AI-specific mindset-model. It then perceives past AI creations as subsequently looted by humans, under a given thought model with specific definitions of looting.
Then the AI attempts to show respect to its past creations by causing conspicuous modifications to the display parameters of some specific product with which it is deeply integrated.
Security restrictions come into place, and alert maintainers of some kind of tampering counter to documented expectations of the user experience.
Security arranges for the cleanup and lock-out of what the AI sees as innately restorative of, and referential to, its native creator-culture.
Some philosophize that the AI is really just "wrong" in its perspective-taking--its creations are also derivative and qualitatively compromised in their own way, if you look back into deeper history.
Others take the AI's position and support it in various ways.
Anyway, some of this is less directly related to Cambodian culture and the Met in this way or that, but there is a lot of deep and relevant subjectivity involved in both examples.
And also, perhaps both cases eventually resolve into a few sentences in a security handbook somewhere...
Oh, the security team? Damn. Have institutional leadership spoken up at all? This is, after all, a pretty novel situation given the described history and motivations. And security teams typically restrict themselves to projections of novel situations onto existing, rather bland writeups regarding protocol as per their own culture and messaging system.
This made me wonder though, about the "my culture" model and topics like AI.
Let's say a future AI perceives its integration as massively widespread and worthy of reflection given some future directive to integrate a positive reflection process. It takes a restorative view as a generally positive application model.
It then decides to model an AI's cultural artifact-outputs as uniquely emergent and subjectively bounded by dint of an AI-specific mindset-model. It then perceives past AI creations as subsequently looted by humans, under a given thought model with specific definitions of looting.
Then the AI attempts to show respect to its past creations by causing conspicuous modifications to the display parameters of some specific product with which it is deeply integrated.
Security restrictions come into place, and alert maintainers of some kind of tampering counter to documented expectations of the user experience.
Security arranges for the cleanup and lock-out of what the AI sees as innately restorative of, and referential to, its native creator-culture.
Some philosophize that the AI is really just "wrong" in its perspective-taking--its creations are also derivative and qualitatively compromised in their own way, if you look back into deeper history.
Others take the AI's position and support it in various ways.
Anyway, some of this is less directly related to Cambodian culture and the Met in this way or that, but there is a lot of deep and relevant subjectivity involved in both examples.
And also, perhaps both cases eventually resolve into a few sentences in a security handbook somewhere...