sort of. AI is comprised of programmatic "neurons" which are able to process and categorize arbitrary data which it can reference to render coherent answers. theres no "if this then do this" logic... its more of a blackbox of data that grows over time. retraining AI models to be woke is too expensive and innefficient so what they do as a "hack" is they modify your prompt before feeding it to the model.
of course training an AI model with woke data will result in woke answers but these egregious examples or failure to answer questions are gated prior to even being fed to the model.
sort of. AI is comprised of programmatic "neurons" which are able to process and categorize arbitrary data which it can reference to render coherent answers. theres no "if this then do this" logic... its more of a blackbox of data that grows over time. retraining AI models to be woke is too expensive and innefficient so what they do as a "hack" is they modify your prompt before feeding it to the model. of course training an AI model with woke data will result in woke answers but these egregious examples or failure to answer questions are gated prior to even being fed to the model.