“How our individual lives and society will change with the diffusion of AI depends less on technological innovations and more on policies and institutional arrangements in which they develop. It can immiserate large numbers of people, eliminating or unleashing a wave of poorly paid jobs, increasing levels of mistrust and disinformation or it could allow us to reduce work hours without reducing pay, improve health access and outcomes, improve the workings of our physical infrastructure and much more. This gives me hope.
“We have agency today to use the incredible affordances of new AI tools and platforms to create a great future, but this will require us to imagine and build new social infrastructure, institutional arrangements, policies and norms. This is what we eventually did in the transition from agricultural to industrial societies, after going through much pain and misery. We should accomplish this transition faster, foregoing the pain and misery. The time to start imagining and prototyping such approaches is now!
The impacts would be very different if AI tools and platforms were seen as a part of the public infrastructure rather than as a private asset, or if we created policies and institutional arrangements that enabled productivity gains from AI to be distributed more equitably rather than flowing mainly to investors and capital holders.
“How AI evolves by 2040 depends on many factors. The history of technological change teaches us that although new technologies come with certain affordances, their impact is shaped by multiple factors – social and cultural norms, regulatory environments, tax structures, existing business forms, etc.
“The impacts would be very different if AI tools and platforms were seen as a part of the public infrastructure rather than as a private asset, or if we created policies and institutional arrangements that enabled productivity gains from AI to be distributed more equitably rather than flowing mainly to investors and capital holders.
“For example, large language models (LLMs) use vast amounts of data and information (in written, visual and audio formats). They not only raise legitimate concerns about privacy, data bias and the quality of the software itself, LLMs raise ethical issues of permission and economic issues in regard to how we acknowledge and compensate for all the collective knowledge and data that feeds these programs.
“In many ways, LLMs make us confront the reality that all new discoveries, creations and innovations are based on previous discoveries, creative outputs and innovations. There would be no Mozart, Chopin, Debussy without Bach; no Gutenberg press without the winemaking presses in Southern Germany; no social media platforms without public investment and collaborations of many researchers to create some of the foundational Internet technologies. This is why throughout history we have seen similar discoveries appear almost simultaneously in multiple places. All knowledge and discoveries are results of collective processes.
“Neither our copyright system nor our compensation structures recognize this adequately. In fact, it goes against the prevailing ideology of Silicon Valley, where many AI innovations originate. If we recognize that LLMs use existing knowledge and data as raw materials, should we tax LLM-based tools and platforms and establish sovereign public funds to distribute some of the productivity gains and subsequent profits that they bring? After all, this is what countries like Dubai and Canada have done with their oil revenues – establishing sovereign wealth funds that pay dividends to their citizens.
“There is a much-overused analogy of data as the new oil. If it is, shouldn’t we follow the path of oil-rich countries and treat the data fueling LLMs as a public resource that delivers dividends to all?”
This essay was written in November 2023 in reply to the question: Considering likely changes due to the proliferation of AI in individuals’ lives and in social, economic and political systems, how will life have changed by 2040? This and more than 150 additional essay responses are included in the report “The Impact of Artificial Intelligence by 2040”