Human Learning and Machine Learning— Challenges and Opportunities for AI in Education

The arrival of gene­ra­ti­ve AI has pre­ci­pi­ta­ted a panic among some educators while prompting qualified enthu­si­asm from others. Ask a GPTs (Generative Pre-Trained Transformer) to take a high-stakes select response test, and it will generally pass or even excel. Prompt it to write a five-paragraph essay on any topic, and it will produce a well-written if dull and pre­dic­ta­ble response, at least as good or better than a student’s response. Here comes a cheating bonanza: unless strictly proctored, there’s no reliable way to tell whether a student response has been generated by AI, espe­ci­al­ly with a few extra tricks in written text such as adding some typos or awkward expres­si­ons. What to do? How to put gene­ra­ti­ve AI to good use?

In addition to the broad range of regular themes addres­sing edu­ca­tio­nal and social change at The Learner Conference, this year we will have a special theme exploring the impli­ca­ti­ons and appli­ca­ti­ons arti­fi­ci­al intel­li­gence in education generally, and GPTs in par­ti­cu­lar. Presentations addres­sing this special theme will examine the dif­fe­ren­ces between machine and human learning and the ways in which machines can com­ple­ment human learning. Some of the questions we will address are: What are super­vi­sed, unsu­per­vi­sed and rein­force­ment machine learning? Will mecha­ni­cal intel­li­gence ever be able to replicate or exceed human intel­li­gence? What are the practical appli­ca­ti­ons of learning analytics and arti­fi­ci­al intel­li­gence in learning manage­ment systems and other edu­ca­tio­nal tools? What are the impli­ca­ti­ons for diversity, equity and inclusion?