Nowadays, the use of generative AI (like ChatGPT) has become common in colleges and universities. Students are using it a lot for their studies and professors are using it for teaching. But it is not just a new technology, it has completely changed the way we learn and take exams.
Our research conducted with 28 professors and experts from Canada has revealed that a huge change has taken place in the field of education. Now the biggest question is that when computer programs (AI) can think and work like humans, then what ability should we test the students? Should we be looking at their memory or how they use this technology?
Research on AI and academic integrity
Research over the last 15 years has made it clear that AI in education is like a double-edged sword, which has its own advantages and disadvantages. On one hand, AI has become so adept at writing human-like language that it has become difficult to detect imitation and it sometimes spreads misinformation or discriminatory things, on the other hand, it is also proving to be a great help for disabled students and those learning new languages. Since it is now impossible to block every AI tool, schools should improve their rules and teach students and teachers how to use this technology appropriately and responsibly instead of focusing solely on catching cheating.
Three major areas of evaluation
The teachers involved in our research believe that their job is not just to catch cheating but to promote honest studies. He lists three main ways where we can see whether students are using their intelligence or not.
Giving correct order (prompt) is also an art
The first is prompting, that is, giving correct instructions to AI. Now this is also being considered as a skill because to get the right work done by AI, the student has to express his views clearly and divide the big work into small parts. But this is correct only when the student does it not secretly but honestly on the basis of his knowledge. The second method is critical thinking. Since AI sometimes conveys false or fabricated things as truth, it is now necessary to teach students not to blindly trust what AI says, but instead look for shortcomings and mistakes in it. In the coming time, it will be needed a lot in offices also.
It is okay to take help of AI in brainstorming but the ideas should be your own.
Most of the debate is about writing. Here the teachers have drawn a clear Lakshman Rekha. It is okay to use AI for brainstorming, provided the ideas are the student’s own. Using AI to improve the grammar (editing) of what one has written is also appropriate when the student has written the original text himself. But getting an entire article or draft written by AI is considered absolutely wrong, because teachers believe that only when the student struggles to write on his own, his mind truly develops.
What is the ‘post-plagiarism’ era?
We have now reached an era where humans and AI are writing together, and because of this the old definition of ‘real written work’ is changing. This has been named ‘post-plagiarism’, which does not mean that it is now okay to steal or be dishonest. Its real meaning is that if a person is creating something new with the help of AI, then it cannot be called direct copying or theft. The need of today is that we teach students to make AI not their enemy but a companion and learn to co-create new things with it.
Five Design Principles for the Future
We have suggested 5 main things (principles) to properly assess the ability of students in the era of AI:
First of all, the rules should be very clear. The teacher should be told in advance whether the help of AI is to be taken in any homework or project or not, and if so, how much. Second, we should look at the way things are done, not just the end result. That is, how the student started, how he made notes and how hard he worked. Third, students should be given tasks that involve the brain more, such as giving an opinion on a serious problem or taking a decision based on the environment around them, which AI cannot do easily.
Fourthly, students need to be taught to recognize the shortcomings of AI, so that they can know where AI is making mistakes or where it is biased. And the last most important thing – to give importance to the student’s own thinking. The test should be about how the student learned something and what his opinion is about it, not just about how much information he remembers.
