Creating Inclusive Digital Economies with Generative AI: Integrating Neural Networks to Expand Accessibility and Equity in Financial Services

Abstract

This essay explores the potential of generative deep learning models—neural networks that operate much as human brains operate—to expand inclusion and access to financial services, and to build inclusive digital economies. Philosophers of technology and data science scholars have explored these models as an age of critical significance. To do so, they inquire deeply into the future of the digital world that is quickly emerging from their seemingly endless capacity to invent not just new words and strings of words but new possibilities about what the words might—could equivocally mean. This is also a future in which, somehow, generative models that usher us into ever more immersive virtual realities play an intensifying part in the experiences of the many, not just the few. However, we are interested in generative AI quite unlike the ones described in these imaginaries. Our focus is neural networks that draw on an archive of data to create what we call "does well enough" equities and asset prices. Our focus is on what might happen if these economic, rather than purely linguistic and virtual, generative models became pervasive. We find they are powerful in opening a previously possible future. This future is one of vernacular sophistication across users; a world in which entities can be creditors and grow—even slide into less debt—but in a society where borrowing or growing financially are but one of many activities and welfare. We speculate below about what this means for inclusive potential, as well as the normal that might follow, inhabit, or supersede our operational inclusivities. Overall, we guess the next big people will be the ones who work with the models to free money from its supposed unique and mysterious value, not just to produce ever more complex instances of data.