ChatGPT: How might the future of homework be affected by the artificial intelligence programme?

ChatGPT can offer convincing human-sounding answers to exam questions. Credit: ChatGPT

The use of artificial intelligence in emerging technologies continues to advance at pace, causing a headache for teachers trying to detect the authenticity of a student's homework.

An online service, called ChatGPT, has sparked alarm in some within the education system due to its ability to generate convincing essays which can easily fool much of the existing anti-plagiarism software.

We take a look at how ChatGPT works and what it means for the future of homework:

What exactly is the ChatGPT software?

ChatGPT, released by the Silicon Valley company OpenAI in November, is a software application made to mimic human-like conversation based on user prompts.

The chatbot model, programmed using a machine learning technique called Reinforcement Learning from Human Feedback, has been trained on a massive amount of text data from the internet to enable it to do this.

It can mimic dialogue, respond to questions, acknowledge mistakes, reject inappropriate requests and push back against flawed premises.

The chatbot can be used for a range of purposes, including digital marketing and online content creation. Some teachers have reportedly said it could even help with marking and writing school reports.

OpenAI, however, has recognised the software's tendency to react with "plausible-sounding but incorrect or nonsensical answers".

"Ideally, the model would ask clarifying questions when the user provided an ambiguous query. Instead, our current models usually guess what the user intended," the company said on its website.

"We’re eager to collect user feedback to aid our ongoing work to improve this system."

You can read more about ChatGPT here.

Who invented the technology?

OpenAI, a research and development firm, was launched in 2015 by Silicon Valley investor Sam Altman and billionaire Elon Musk, the SpaceX, Tesla and Twitter boss.

Mr Musk, who has described ChatGPT as "scary good", left OpenAI in 2018.

Mr Altman declared on Twitter in early December that ChatGPT had exceeded one million users.

Examples of the ChatGPT software in action?

One user, for example, fed ChatGPT an English exam question and the software responded with a five paragraph essay about Wuthering Heights, the classic 1847 novel by Emily Brontë.

Another user asked the chatbot for an essay about the life of William Shakespeare four times using the same prompt on each occasion. He received a unique version each time.

Can the software help student's cheat?

Anyone who registers for an account can type questions into it, meaning the chatbot's algorithms can answer requests for essays on any subject.

Amid fears the tool is making it easier for students to cheat on homework, New York City public schools are to ban students and teachers from using ChatGPT.

“Due to concerns about negative impacts on student learning, and concerns regarding the safety and accuracy of content, access to ChatGPT is restricted on New York City Public Schools’ networks and devices,” Jenna Lyle, the deputy press secretary for the New York public schools, said in a statement.

“While the tool may be able to provide quick and easy answers to questions, it does not build critical-thinking and problem-solving skills, which are essential for academic and lifelong success.”

Are there concerns around ChatGPT's use in Britain?

Schools in Britain, Ofqual, the exams regulator for England, and the Department for Education are reportedly monitoring whether the programme is being used by students to cheat.

Geoff Barton, general secretary of the Association of School and College Leaders, is quoted as saying: "We're aware of concerns that ChatGPT may be used by some pupils to write answers and pass this off as their own work, although we have not directly received reports of problems from our members thus far.

"We will be keeping a close eye on how this develops and, if this does become a problem, we will certainly be pressing the government to provide guidance and support."

Ofqual has considered whether schools should be given guidance on how to stop students from cheating on coursework with the chatbot, according to the Daily Telegraph.

An education department spokesperson has said: "We are aware of reports from school leaders around pupils' use of artificial intelligence to produce written work.

"Schools and teachers know their pupils best and are experienced in identifying individual pupils' work.

"There are strict rules in place to ensure pupils' work is their own. The department has regular, routine engagement with Ofqual, exam boards and school leaders to ensure the fair and effective running of exams and qualifications and will continue to do so in the run-up to exams this summer."

ChatGPT's profile grew in Britain after Luke Evans, the MP for Bosworth, told the Commons: “I think this may be a first”, as he shared a speech compiled by the AI programme, which had been given the command to write a Churchillian speech on the state of the UK over the last 12 months.

Want a quick and expert briefing on the biggest news stories? Listen to our latest podcasts to find out What You Need To Know