1. What are the implications of students using Generative AI to copy and paste essays which are then evaluated by AI systems, meaning that humans are not involved in the process of education other than as ‘users’ of an interface?

 2. What might happen to creativity once all sources Generative AI draws upon are themselves created by Generative AI? These are open questions with no right or wrong answers, the world does not run on one set of rules.

'What do you base your understanding on? Do you refer to the adverts which come from technology companies, your own experience and/ or evidenced based research which does not have a vested interest in selling technologies?'

'Where can you find credible evidence to support your ideas and experiences? In the lecture there are a number of links you can follow up to find out how researchers analyse these themes and justify their conclusions'

The second problem, which I think comes up with the first, is creativity homogenization. Experiments conducted by the University of London and Exeter University have shown that when human writers use AI-generated ideas for five stories, the originality and practicality of their works increase by 26.6%, but the similarity between the works of different authors increases by 10.7%. Researchers warn that if AI-generated content becomes widely adopted in the publishing industry, collective innovation may dwindle due to the homogenization of content. This is because AI relies on probability models to reorganize existing data, lacking the cross-domain associations and emotionally-driven innovation that are unique to humans.

The second is reduced creativity. According to research published in a Nature journal, if the training data of a new generation of AI models consists only of content generated by previous models, errors and biases will accumulate with each generation. For example, the error rate of the third generation model in the GPT series on mathematical reasoning tasks is 17% higher than that of the second generation, because the training data contains the error output of the previous generation. Like genetic defects caused by biological inbreeding.

2. When AI only learns from other AI's work, creativity could become like photocopy of a photocopy - each version getting worse. Like how memes get distorted when shared too much.

3. I base this on normal student life - seeing classmates use translation apps for English homework, or how TikTok trends repeat until they're meaningless. Tech ads always promise magic solutions, but real life is messier.

4. For instances ,we can look at everyday results: students who overuse translation tools often can't speak properly, or how Instagram filters create unrealistic beauty standards. The lecture links probably show similar patterns in academic research.

2. If Generative AI only uses content made by other AI, creativity might become very limited. The same ideas could repeat over and over, with no new or unique thoughts. My understanding comes from a mix of things: my own experiences with AI tools, like using them for homework help; and some simple research I’ve read, like articles from schools that talk about AI’s good and bad sides.

To find credible evidence, I can look at university websites or research papers—they don’t try to sell products, so their ideas are more honest. The lecture links to such studies, which help explain how AI affects creativity and learning in a real way.