generative experiences
there is a lot of controversy around how generative ai will impact creative disciplines, especially media. the controversy revolves around whether ai generated "art" is really art, or should be valued to the same degree as art that humans create them. like most things online, both ends of the spectrum are pretty extreme and lack nuance. my overall position is that "art" is meant to express meaning about the human experience in a profound and unique way. if generative ai can do that, then it should be valued; however, i have not seen a lot of generative ai content that produces and sense of introspection, though the failures of generative ai (comical images, memes, and videos) have produced some joy.
we are still at the very beginning of understanding how generative ai will shape the media and content we consume. mark zuckerberg recently talked about how we are at the beginning of a third era in social feed content. the first was friends and personal connections, the second was influencers and creators, and the third will be ai generated content. i think consuming a feed of exclusively ai generate content is a little dystopian, because inevitably coupling generative ai content with the mechanisms of virality will lead to an echo chamber effect. gen ai algorithms will be incentivized to produce content that people will engage the most with, and we've learned a lot over the last few years that the content people engage the most with can be inflammatory, full of lies, and harmful.
i am still optimistic, primarily because new technology always introduces positive and negative externalities. the same will be true for generative ai content. successful media platforms will have create high quality and safe feeds for consumers in the long run. so while it is important to guard against negative consequences, i am also excited about how generative ai will allow people to create new kinds of media and experiences. the internet exponentiated human creativity, and gen ai is the next iteration.
the key piece that a lot of current generative ai art misses is we aren't really producing anything new. you can create stories, photos, and videos using foundation models, but ultimately this content is a copy of what we already create and consume. some of it lives in the uncanny valley of realism; for generated content that surpasses that barrier, the primary excitement comes from the novelty that a computer generated it, not that the content itself is particularly unique or profound. i think that generative art for consumers will gain real, tangible traction when foundation models are used to produce something new and novel that is fun for people to engage with on a repeated basis.
using the internet is an experience. when you open xitter, instragram, or tiktok- you are accessing a social experience that delivers content to you for engagement. posts, likes, reactions, and comments are the modes of interaction in the experience. right now, that experience is curated by a set of machine learning recommendation algorithms that use your previous interactions within the experience to dictate the rest of the experience. so right now, your experience using the internet is shaped by your interactions with it, albeit in a restricted sense.
my thesis on consumer generative ai is that foundation models will supercharge how digital experiences change dynamically based off our interactions with them. rather than being delivered as a rigid experience, my interactions with technology will be shaped by foundation models that construct novel experiences just for me. generative ui's, language oriented actions, and an expanded personal digital identity that includes the context of my needs, desires, and preferences will create personalized and federated experiences unique to me. different applications will unify into one interface that materializes based off what i need to accomplish. a particularly exciting aspect of this is how these experiences will "age" with us, as we go from child to teenager to adult.
so we will all have federated, personalized digital experiences, but things will get even more interesting when you overlay these digital experiences over the internet's emergent social network. after all, one of the internet's first "killer apps" was forums. what is the generative ai equivalent of a forum, a digital third place where people come together to interact, share ideas, and have fun? what will these digital spaces look like with thousands and millions of ai agents, created by humans to engage with the broader world in different ways? how will these spaces be shaped, what functionality will they have, and how will people impress themselves upon them when the power of their digital identity seamlessly integrates with foundation models? how do we create spaces and experiences with generative ai that further human connection and experience? these are daunting and awe-inspiring questions to explore, and i am excited to see how it plays out as technologists begin to shape the next iteration of the internet.