Yet 56% of education leaders are concerned about the ethical implications of AI, including privacy risks and bias in algorithms.

This webinar discusses the limitations of generative AI, such as ChatGPT, for high-stakes grading and proposes classification AI as a superior alternative. We define generative AI as "the world's greatest autopredict algorithm" and explain its mechanism for tokenising input and predicting the next most likely token. It then delves into the problems associated with using generative AI for grading.

Key points include:

• Generative AI Explained: It's described as a sophisticated auto-predict algorithm that tokenises input and predicts the next most likely token.
• Problems with Generative AI in Grading: The document highlights issues such as bias, inconsistency, and lack of explainability.
• Academic Findings on Generative AI Bias and Inaccuracy
• Classification AI as the Solution

Watch the full replay!