Many people are worried that teachers’ jobs will be in danger as AI takes over with time. However, these fears are misplaced. Harvard University is making strides in integrating artificial intelligence (AI) into its coding program. The university plans to employ an AI chatbot with the capabilities of ChatGPT as an instructor in its renowned Computer Science 50: Introduction to Computer Science (CS50) course.

Based on OpenAIs GPT models, the chatbot will offer feedback and help students with their coding work. The project could prove a significant milestone in the evolution of AI tools for education.

However, using AI in a college class has also sparked concerns about its ethical and security risks. Many educators fear that the technology will enable students to cheat and plagiarize in an undetectable way by human teachers. Others worry about security risks, such as the potential for tampering with student data by malicious actors. The move by Harvard will likely lead to increased scrutiny of AI technologies for educational purposes, especially those that involve direct interaction between humans and computers.

Despite these worries, the AI chatbot is a promising innovation that can help students with their coding work and provide a valuable resource for the school. It will be able to answer questions and assist students with writing code quickly and efficiently. It can provide a more personal experience for students by understanding their learning styles.

• RELATED STORY: H&M Shares Jump 11% As Summer Collection Boosts Profit

The AI in question, called LUCI, has been developed by a team of Harvard graduate students and professors. It uses a deep learning model to understand the intent of a user’s query and then delivers an accurate response. It can also identify grammatical errors and help with spelling. Moreover, it can quickly answer complex questions and provides students with the information they need to complete assignments.

LUCI is trained on thousands of real-world conversations with students enrolled in the CS50 course to improve its performance. Students can rate their answers and provide feedback to make the bot more effective. The results of this training will then be used to train a new generation of AI chatbots.

What’s Inside

Despite the many inevitably arbitrary engineering choices that went into creating ChatGPT, the underlying technology is relatively simple. The system’s core is a neural net that can perform computationally irreducible computations. That means that the machine can, in principle, figure out anything a human can tell about the world and its nature.

To generate text that’s “like” what it’s read, the network learns with the help of a large dataset of human-created web content, books, etc. Once the neural net has finished this basic training, it can start from a prompt and continue producing text that’s “like” what it’s been shown. It’s like learning to play a musical instrument by passively listening to music and watching musicians.