Episode Details

Welcome to Revise and Resubmit, where we dig deep into academic papers and explore ideas that challenge the way we think about the world. Today, we’re tackling two fascinating papers that explore a concept we all know, but may not fully understand: bullshit.

In the first paper, "ChatGPT is Bullshit," the authors Michael Townsen Hicks, James Humphries, and Joe Slater argue that the term “AI hallucinations” doesn’t quite capture the falsehoods produced by large language models like ChatGPT. Instead, they propose that we call it what it really is: bullshit. But not just any kind—bullshit that’s indifferent to the truth, like a person who talks without caring whether what they say is accurate.

The second paper, "Bullshit and Organization Studies," by Lars Thøger Christensen, Dan Kärreman, and Andreas Rasche, takes this idea even further into the business world. They argue that bullshit isn't just harmful—it’s sometimes necessary. In organizations, bullshit can serve a performative role, especially in management, where commanding and strategizing might rely on a healthy dose of it to keep things moving.

So, is bullshit a problem, or a hidden tool we’ve been using all along? And when it comes to AI, what does it mean if the machines we create have become the ultimate bullshitters? Let’s dive in. But here’s a question to think about: If bullshit is inevitable, should we learn to master it or call it out for what it is?

References

Hicks, M.T., Humphries, J. & Slater, J. (2024). ChatGPT is bullshit. Ethics and Information Technology, 26(38). https://doi.org/10.1007/s10676-024-09775-5

Christensen, L. T., Kärreman, D., & Rasche, A. (2019). Bullshit and Organization Studies. Organization Studies, 40(10), 1587-1600. https://doi.org/10.1177/0170840618820072