Seton Hall University welcomed alumnus Joshua Meyer ’12 to a workshop entitled “AI 101” hosted by the Teaching, Learning and Technology Roundtable (TLTR) Artificial Intelligence Committee on November 12 at the Bishop Dougherty University Center. This session brought together students, faculty, and administrators from across campus interested in understanding how the latest artificial intelligence tools work and how they can be used responsibly in academic and professional settings.
TLTR AI Committee Co-Chair Jessica Rauschberg opened the program by introducing the committee’s goals and TLTR’s broader mission to support thoughtful, ethical, and effective use of technology. Rauchberg then introduced Meyer, an AI researcher and industry leader whose work spans language technology, machine learning, and product development. Meyer holds a Ph.D. He holds a BA in Automatic Speech Recognition from the University of Arizona and a BA in Liberal Arts from Seton Hall.
Meyer began by explaining the basics of large-scale language models (LLMs), explaining that LLMs are systems trained to predict the next word in a sequence, rather than traditional databases that store and retrieve facts. He demonstrated that while this predictive design enabled the LLM to produce fluent text, it also made it more error-prone when attempting tasks such as mathematics and detailed fact recall. Meyer gave an example of these limitations, including solving a simple arithmetic problem.
Meyer then introduced the concept of agent AI. Agent systems can call external tools or perform actions to complete tasks. Using the example of a model that triggers a calculator program on a computer that calculates 646 plus 101, Meyer explained how these agents can extend the capabilities of LLM by connecting them to tools that can perform tasks that they cannot reliably handle internally. He emphasized that this change represents a new frontier in AI development as it connects inference models with real-world functions and decision-making.
He also mentioned common problems such as hallucinations, where the model produces confident but incorrect statements. Meyer shared examples that included fabricated biographical details about himself, including that he attended Seton Hall, played football in the NFL and appeared on NBC’s “The Office.” He shows how these systems can create plausible but erroneous information when presented with questions that lack sufficient context. He reminded attendees that LLM does not understand truth in the human sense and relies only on statistical patterns in language.
Meyer also discussed the growing problem of what he called “AI slop.” This refers to content that appears polished at first glance, but upon closer inspection reveals inaccuracies, vague claims, or clear signs of machine generation. He urged students not to rely solely on AI to produce academic results and advised faculty to remain vigilant about the quality of writing produced by automated tools. He said AI should not replace human judgment and creativity, and that it is essential that these tools maintain “taste” and critical evaluation even as they become more sophisticated.
Throughout his presentation, Meyer emphasized the importance of human control and intentionality in the use of AI. He advised participants to “be a pilot, not a passenger,” and encouraged users to take a controlled approach by providing clear instructions, context, and constraints. He explained that LLMs perform best when given specific instructions and detailed context. A larger context window improves performance and allows you to process more text at the same time.
Meyer concluded by outlining practical ways that faculty, students, and administrators can leverage AI to accelerate research, organize information, and support creative problem solving. He emphasized that while AI can help process large amounts of text and identify patterns, users still need to verify the results and ensure accuracy.
The TLTR AI Committee, co-chaired by Rauchberg and Ruchin Kansal, investigates the impact of AI on higher education, supports policy development, selects best practices for integration, and explores implications for the future of the workplace.
Click here for more information.
Category: Science and Technology
