Artificial intelligence (AI) is rapidly becoming a common feature in school classrooms and adult education worldwide, with many educators and instructors exploring the potential of this technology to improve teaching and learning outcomes. While AI could offer many exciting possibilities, we must be aware of the realities, risks, and ethical considerations that come with powerful models like GPT-3 (the model behind ChatGPT).
This is the first in a series of articles exploring the impact that conversational AI is having and will have on teaching and learning. In this article, I will specifically cover the categories of chat-based learning and conversational AI very broadly and why GPT-3, very narrowly, is creating such a stir.
What is GPT-3?
GPT-3 (short for Generative Pre-trained Transformer 3) is a large language mode (LLM) with an available toolset of artificial intelligence capabilities developed by OpenAI. GPT-3 can generate human-like text in response to specific scenarios and requests or when exposed to key information (such as course materials for a lesson). GPT-3 be fine-tuned for various language tasks, including summarisation, question answering, and lesson planning. For better or worse, GPT-3 has a scary ability to generate high-quality text that is difficult to distinguish from text written by a human. GPT-3 also has an enormous capacity, with 175 billion parameters, which allows it to perform well on a wide range of tasks to assist humans.
GPT-3 has the potential to improve teaching and learning in several ways. For example, at Noodle Factory, we leverage GPT-3 in our “Walter” platform to generate summarised knowledge bases of the content instructors use to teach a class. Additionally, we can generate personalised lesson plans. Also, we provide real-time feedback to students as they work through problems, complete assignments, and take practice exams. GPT-3 can also generate explanations or summaries of complex concepts, making them more easily understandable for students.
Impact of ChatGPT on education
I now get daily questions about the impact that ChatGPT will have on education. I welcome these questions because, at Noodle Factory, we have devoted the last several years to developing conversational AI systems for education (in both schools and corporations). In fact, our “Walter” platform utilises several GPT-3 capabilities via API (Application Programming Interfaces) connections with the model.
One of the critical features of conversational AI that sets it apart from other technologies is the ability to generate human-like text and perform a wide range of language-related tasks. These human-like capabilities make it particularly useful for generating personalised lesson plans, providing real-time feedback to students, and generating explanations or summaries of complex concepts. However, it is vital to be aware that AI models can sometimes perpetuate or amplify biases or inaccurate information that may be present in the data or language models they are trained on. Instructional designers, teachers, and students must carefully and critically consider the data used to train these models.
In addition to the potentially amazing benefits of AI in classrooms and workshops, there are also risks that we should be aware of. One of the main risks is the potential for the misuse of these tools or for students to rely too heavily on them rather than developing their own knowledge, skills, critical thinking, and problem-solving abilities. It is also essential to consider the impact that the increasing use of AI in education could have on the workforce, as there is a risk that it could lead to the loss of jobs in education, particularly in lower-skilled roles.
Privacy is another key concern when it comes to the use of AI in education. Some AI tools may access substantial amounts of personal data, and it is crucial to ensure that appropriate safeguards are in place to protect student privacy. Educators should also be transparent about the data being collected and how it is used to avoid the “black box problem” in which no one is quite sure how the AI works.
In the next instalment of this series, we will consider how conversational AI catalyses the type of student learning outcomes and impact educational luminaries such as Skinner and Bloom envisioned more than 50 years ago.
--
We hope you enjoyed our series about GPT-3! If you did, make sure to mention us and share the post on Facebook, LinkedIn, and Twitter.
Read the second post of this series here: GPT-3 and Me: How Chat and Conversational AI Are Changing Education (Part II)
We are also excited to announce that we will be adding GPT-3 capabilities to our award-winning AI teaching platform, Walter. Join our waiting list to gain early access when it's launched!