Have you started to incorporate ChatGPT or other generative AI tools into your daily life?

If so, you know the trade-off well: either accept the AI-generated output as “good enough” or put time—sometimes a lot of time—into personalizing it so that it faithfully reflects your unique style or perspective. Kellogg’s Sébastien Martin calls this AI’s efficiency–fidelity trade-off.

This week, we’ll ask: What does this trade-off look like when everyone uses the same AI tools? We’ll also discuss hiring problems facing the nonprofit sector.

Societal impact

As generative AI tools become increasingly popular throughout the population, what changes can we expect to see in the content we collectively produce? And how should managers be thinking about this?

Martin, an assistant professor of managerial economics and decision sciences, teamed up with colleagues Francisco Castro and Jian Gao to investigate this question. The researchers built a mathematical model to simulate how individuals with different preferences might choose to use the technology and what those choices mean in aggregate. Here are four takeaways from their study:

1. Using AI to create content will increase the homogeneity of what we collectively produce, even if we try to personalize the output.

2. This content will also inherit any biases the AI may have acquired during its training process. In other words, the tastes and biases of a few AI company workers may ultimately permeate throughout society.

3. The problems of homogeneity and bias will be compounded as AI-generated output is used to train the next generation of AI—what the researchers call a “death spiral.”

4. On a more positive note, however, the study suggests that creating interactive AI tools that encourage user input and facilitate manual edits can prevent the worst of these outcomes. That’s because interactivity will help AI systems learn to better represent the broad range of styles and perspectives present in the population.

While managers might not concern themselves with these societal-level impacts, understanding the fundamental trade-off between efficiency and fidelity will be key to maintaining the quality of what is produced. Says Martin, “If you put a lot of pressure, time pressure, on your employees, they will use the AI extremely differently than if you give them more time to be creative.”

You can read more about this study in Kellogg Insight.

Talking point

Nonprofits are struggling right now to hire and retain talented staff, according to Kellogg clinical professor and executive director of Center for Nonprofit Management Liz Livingston Howard. Existing employees are burned out after the pandemic, and recruiting new ones is challenging due to the in-person nature of many direct services roles. Inflation is also taking a bite out of budgets at a time when staff are negotiating for better salaries and benefits.

You can read more in Wallethub.