News — On this episode of Business and Society, and faculty director of the initiative returns to the podcast to discuss his research. He covers the impact generative AI has on scientific publishing and the ways databases of corporate annual reports distort data to create an inadequate picture of today’s economy.
The discussion closes with a plea for recognizing the importance of well-funded, nonpartisan, faithful information to better understand the economic world.
Scientific Publishing and Generative AI
Open-access publishing idealistically makes research free and accessible to the public. However, it creates a business model that allows journals to publish endless amounts of work while charging authors a fee. Now, with generative AI so readily available, articles are easier to write, edit, and publish in bulk, and it becomes difficult to assess the quality of any given study.
Professor Davis stresses that the end goal of science isn’t to publish PDFs; the point of science is to engage in a conversation so that science can continuously improve. This, he states, is not something to hand over to machines.
“If a couple of people and a bot can write 288 presumably publishable papers, we’re just going to be surrounded by a sea of publications.”
Distortion and Inaccuracy in Corporate Annual Reporting
Companies continue to disclose information in 10Ks as laid out in the Securities Exchange Act of 1934. At the time of creation, these filings were primarily focused on tangible assets, but today, most company value is usually intangible. Corporate annual reports continue to measure aspects of companies that are not as relevant when looking at today’s economy. However, people still rely on these reports when trying to understand what the modern economy looks like, resulting in a distorted picture of reality.
“We’re still largely stuck with the kind of information that made sense during the Great Depression. We created a set of requirements for companies during the 30s and early 40s, and we’re still living with the consequences.”
The Value of Faithful, Honest, Nonpartisan Data
Responses to government surveys have been declining, and so has funding for such surveys. Davis states that thoughtfully created government-mandated surveys are necessary to gather good data about the way the world works. Without honest, nonpartisan data about the economy, it is impossible to operate effectively within it.
RSS