Reflecting on your question "if the medium is the message, what is the message of generative AI," I think the message of generative AI is personalization. If someone is getting information from a chatbot, depending on the specific words they use for their prompt (or the user's chat history), they will get a slightly different output than someone else. This probably results in increased fragmentation of experience / understanding. Interestingly, GenAI even enables people to personalize the medium itself - to transform any input article, diagram, etc. into a new medium that better matches their learning style (like video or audio) - at least, this is theoretically possible once the technology advances further. So, perhaps the influence of the "source medium" on a society's collective perception of the event becomes less important, and "the medium is the message" as a concept becomes relevant mostly on an individual level. It could increase society's collective understanding, by unlocking certain content previously locked in one medium inaccessible to some (e.g., a visual diagram) into a medium accessible to those individuals (e.g., audio), or vice versa.
Another message of this current generation of GenAI could be efficiency over accuracy. People could end up with slightly inaccurate perceptions of events / concepts, unless future generations of the technology iron out the confabulation issue.
Reflecting on your question "if the medium is the message, what is the message of generative AI," I think the message of generative AI is personalization. If someone is getting information from a chatbot, depending on the specific words they use for their prompt (or the user's chat history), they will get a slightly different output than someone else. This probably results in increased fragmentation of experience / understanding. Interestingly, GenAI even enables people to personalize the medium itself - to transform any input article, diagram, etc. into a new medium that better matches their learning style (like video or audio) - at least, this is theoretically possible once the technology advances further. So, perhaps the influence of the "source medium" on a society's collective perception of the event becomes less important, and "the medium is the message" as a concept becomes relevant mostly on an individual level. It could increase society's collective understanding, by unlocking certain content previously locked in one medium inaccessible to some (e.g., a visual diagram) into a medium accessible to those individuals (e.g., audio), or vice versa.
Another message of this current generation of GenAI could be efficiency over accuracy. People could end up with slightly inaccurate perceptions of events / concepts, unless future generations of the technology iron out the confabulation issue.