Opinion: Why We Should Mandate AI Literacy Courses

"AI Literacy" [from ScienceDirect] is defined as “the ability to understand the basic techniques and concepts behind AI (artificial intelligence) in different products and services.”

On May 22, 2007, "Charlie Bit Me!"—a wholesome and hilarious video of a toddler named Charlie biting his brother—was posted to YouTube and quickly amassed over 92 million views to become the most viewed video on the platform. Almost two decades later, the video has experienced a second surge in virality: in a trending version of the video, a strange man in a Kylo Ren mask interrupts the iconic moment, with ominous music playing in the background. This video and others like it, part of the mysterious "time traveler interrupts" meme trend, have left viewers disturbed and confused. Recent insight reveals that the trend is fully AI-generated, blending original footage with synthetic elements. The lack of awareness of the implications of this trend and the power of generative artificial intelligence highlights an urgent issue: the need for AI literacy in our society. As AI-generated material becomes increasingly indistinguishable from unaltered content, understanding its implications is integral to maintaining media integrity, protecting users from malicious content, and ensuring responsible media consumption.

Artificial intelligence literacy has been relevant for decades, but—following recent developments in generative AI and increased consumer usage—AI-centric education has become more crucial than ever. Artificial intelligence literacy is defined as "the ability to understand, use, monitor, and critically reflect on AI applications" (Laupichler et al., 2022). GenAI's ability to create convincing yet entirely falsified content poses a new threat to media integrity. Recently, there have been reports of Midjourney, a generative AI bot, producing photorealistic fake images of historical events, demonstrating AI's rapidly advancing capability to fool users.

AI's ability to generate persuasive content creates opportunities for profit in digital markets and corresponding risks of misinformation and malice (Goldfarb & Tucker, 2023). A Haas research project I’m working on demonstrated through A/B testing that AI-generated ads outperform human-created ones in click-through rates, creating positive network externalities as more advertisers adopt AI tools. Unfortunately, this increased efficiency is not completely indicative of beneficial productivity—in fact, 450,000 malicious programs are registered daily (AV-Test, 2024). AI's potential to create appealing content opens the door to significant misuse. AI-generated ads could spread malware and misinformation, eroding public trust in legitimate sources (Ferrara, 2024).

As AI integrates into various aspects of our lives—from political rhetoric to healthcare—understanding AI-aided content becomes a fundamental form of human capital (Acemoglu & Restrepo, 2023). Reports suggest that AI and automation could displace 85 million jobs by 2025, creating 97 million new roles focused on AI system regulation (World Economic Forum, 2020). The opportunity cost of not incorporating AI literacy into educational programs is substantial: decreased employability, increased vulnerability to scams, and reduced technological adaptability. Making AI literacy a required course for all students represents an investment in human capital that can yield significant returns through enhanced job prospects and better protection against AI-enabled threats.

Criticism of AI-centric education centers around the funding and effort needed for nationwide programs. However, when considering the economic principle of comparative advantage (Autor, 2022), AI literacy is essential to maintaining workforce competitiveness. Schools have already incorporated lessons about online threats, so adapting programs to include AI literacy is reasonable. Berkeley already offers AI ethics courses and programs that specifically teach about the dangers of AI. These programs don't always require huge funding; AI itself could help propose cost-effective literacy programs for students, workers, and the elderly.

While obvious AI-generated content may be easy to identify, more people are falling prey to sophisticated AI scams and misinformation. Cybercrime is rising, and AI augments the ease of malicious content diffusion. A general understanding of AI's capabilities and limitations is imperative to navigating the online world and must be incorporated into fundamental education. As AI becomes more sophisticated and the line between synthetic and real content blurs, we must educate ourselves and remain vigilant or risk being misled and manipulated.

Works Cited

AV-Test. (2024). Malware statistics & trends report. https://www.av-test.org/en/statistics/malware/

Ferrara, E. (2024). GenAI against humanity: Nefarious applications of generative artificial intelligence and large language models. Journal of Computational Social Science. https://link.springer.com/article/10.1007/s42001-024-00250-1

Laupichler, M. C., et al. (2022). Artificial intelligence literacy in higher and adult education: A scoping literature review. Computers and Education: Artificial Intelligence. Elsevier. https://www.sciencedirect.com/science/article/pii/S2666920X2200056X

World Economic Forum. (2020). The future of jobs report 2020. https://www.weforum.org/reports/the-future-of-jobs-report-2020

Xiang, C. (2023, March 27). People are creating records of fake historical events using AI. VICE. https://www.vice.com/en/article/k7zqdw/people-are-creating-records-of-fake-historical-events-using-ai

Goldfarb, Avi, and Catherine Tucker. "Digital Economics." Journal of Economic Literature, vol. 61, no. 1, 2023, pp. 3-54.

Autor, David. "The Labor Market Impacts of Technological Change: From Unbridled Enthusiasm to Qualified Optimism to Vast Uncertainty." NBER Working Paper Series, National Bureau of Economic Research, 2022.

Acemoglu, Daron, and Pascual Restrepo. "Tasks, Automation, and the Rise in US Wage Inequality." Econometrica, vol. 91, no. 1, 2023, pp. 1-45.