LeBlanc, along with his associate at SNHU, George Siemens, is pioneering a novel AI solution, termed Human Systems, which addresses various facets of university life. Their objective is to leverage technology for positive impact while sidestepping common pitfalls associated with it.
This solution is slated for release in the summer of 2024.
During the IFE Conference hosted at Tec de Monterrey in Mexico, participants voiced concerns that algorithmic bias could disproportionately affect the Global South, highlighting it as a major issue in the realm of AI.
"Every dataset carries inherent biases. Therefore, the pivotal question is: how do we prevent detrimental bias?" Siemens posed to the audience.
He continued, "When working with data or constructing models, decisions are made regarding inclusion and prioritization of data points. For instance, employing a predictive-next-word token approach inherently incorporates biases from its creators."
Michael Fung, the executive director of Tecnologico de Monterrey’s Institute for the Future of Education, highlighted the growing awareness of bias discrepancies between the global north and south, catalyzed by the proliferation of AI in higher education.
"While complete eradication of bias might be unattainable, it's imperative to acknowledge and elucidate the assumptions and constraints associated with data," remarked Fung.
LeBlanc criticized the higher education sector's inadequacy in managing data effectively, stressing the need for a global data consortium to enhance AI-driven educational insights. Initially funded by the Bill and Melinda Gates Foundation, this consortium aims to combat algorithmic bias and cultural dominance, particularly concerning LLMs with limited sample sizes.
"We must transparently identify the data sources informing AI outputs to comprehend and mitigate biases effectively," echoed Fung.
He added, "Further research is indispensable to refine our practices in this domain."
