As generative AI technologies disseminate, many questions are raised regarding the ethics pertaining AI alignment, the consequence of releasing this technology on an unprepared general public, and social implications of massive data models and the implicit biases baked into the data set.
With the stage set for this technology to change the way we engage with our computers and each other, there are also many concerns pertaining to the concentration of this power in the hands of a few capitalist mega corps, and the guard rails they decide to place on the kind of content their technology generates.
A study in pipelining as many bleeding edge AI models as possible, ShadowSelfie™ forces the observer to subject themselves to verbal abuse at the hands of an inanimate object, taking an objective CV image analysis and applying a prompted GPT call to critique, insult, and interrogate the mirror’s observer, which is read back with Text To Speech technology. Although the observer doesn’t exist, does the blow still land?
