OpenAI claims that teachers will never be able to detect students using ChatGPT because AI detectors are ineffective. The company advises teachers to use ChatGPT as an educational tool to stimulate critical thinking. ChatGPT is an AI tool that generates text based on a conversation. It uses a language model called GPT-3 (GPT-4 in the paid version), developed by OpenAI, which is capable of producing coherent and diverse texts on any topic. However, ChatGPT also raises ethical and pedagogical concerns. Some students tend to rely on the chatbot without learning or thinking. Unfortunately, there is no reliable way to detect if a text was generated by ChatGPT or a human. Tools claiming to do so are ineffective and can make mistakes, according to OpenAI.
Is ChatGPT undetectable by teachers? That’s what OpenAI claims
OpenAI has released a handbook for teachers to educate them about the risks and opportunities associated with ChatGPT. The company advises teachers not to rely on AI detectors but to use ChatGPT as an educational tool to stimulate critical thinking and problem-solving skills in students. According to OpenAI, ChatGPT does not have knowledge of what could be generated by AI and sometimes invents random responses to questions like “did you write this [essay]?” or “could this have been written by AI?” These responses have no basis in reality. OpenAI provides examples where ChatGPT claimed works of Shakespeare or the Declaration of Independence were written by AI. Even if AI detectors improve their accuracy, students could still make slight modifications to their content to avoid detection. This adaptability poses a significant challenge for educators and institutions trying to combat AI-generated academic work.
Using ChatGPT as an Educational Tool
For teachers, ChatGPT can be both a blessing and a curse. However, it is important to use it wisely and responsibly in order to maximize its benefits and minimize its drawbacks. OpenAI believes that ChatGPT can make classroom teaching more interactive and enjoyable. For instance, teachers can ask students to interact with ChatGPT on a specific topic and then assess the quality and relevance of the information obtained. They can also evaluate students’ understanding of the limitations and biases of AI systems. OpenAI states that students can demonstrate their ability to engage with AI and their comprehension of AI system limitations. Educators can assess the quality of questions asked, the relevance of the information obtained, and students’ ability to question, verify, and consider potential biases in the information. Meanwhile, it is worth noting that OpenAI is facing legal action for GDPR violations in Europe. Source: OpenAI (1), (2).