ENG1503
PORTFOLIO Semester 2 2025
Due Date: 23 October 2025
Detailed solutions, explanations, workings
and references.
+27 81 278 3372
, QUESTION ONE – ARGUMENTATIVE ESSAY (2 ANSWERS PROVIDED)
The growing use of Generative AI (GenAI) chatbots in education, workplaces, and
everyday life has raised concern about how such tools affect human thinking. While
GenAI offers convenience and productivity, I strongly believe that excessive reliance
on these tools weakens users’ ability to think critically. This decline in independent
thought is not just about short-term laziness, it presents long-term risks to problem-
solving, reasoning, and decision-making.
Firstly, GenAI encourages cognitive offloading. According to Gerlich (2025),
cognitive offloading happens when people rely on external tools to do the mental
work for them. GenAI chatbots, with their fast answers and summarised content,
make it easy for users to skip thinking through a problem or evaluating information
deeply. As users get used to the convenience of asking a chatbot rather than
reasoning through a topic, they lose the habit of analysing, comparing, and
questioning different sources of knowledge. Over time, this can lead to shallower
learning and a lack of internalised knowledge, especially among younger users
(Gerlich, 2025).
Secondly, Sarkar et al. (2025) show that GenAI changes the way users engage with
content. In their study, knowledge workers admitted that GenAI made their tasks feel
easier, but it also lowered the effort they put into analysing and evaluating the
information. While these tools can support task completion, they also reduce the
amount of mental effort needed, especially in low-risk tasks. When people trust the
AI’s output without checking its accuracy or thinking through its suggestions, their
ability to problem-solve and think independently is at risk. Sarkar et al. (2025) also
found that over time, users may lose the motivation to challenge or improve on AI-
generated outputs, especially when time pressure and job demands make speed
more valuable than accuracy.
Thirdly, Fügener et al. (2021) highlight the issue of poor delegation between humans
and AI due to a lack of metaknowledge, people often do not know when they should
rely on AI or when to think for themselves. In their experiments, humans performed
worse when asked to delegate tasks to AI, not because they disliked AI, but because
they could not judge their own abilities. This overconfidence in AI and
Varsity Cube 2025 +27 81 278 3372
PORTFOLIO Semester 2 2025
Due Date: 23 October 2025
Detailed solutions, explanations, workings
and references.
+27 81 278 3372
, QUESTION ONE – ARGUMENTATIVE ESSAY (2 ANSWERS PROVIDED)
The growing use of Generative AI (GenAI) chatbots in education, workplaces, and
everyday life has raised concern about how such tools affect human thinking. While
GenAI offers convenience and productivity, I strongly believe that excessive reliance
on these tools weakens users’ ability to think critically. This decline in independent
thought is not just about short-term laziness, it presents long-term risks to problem-
solving, reasoning, and decision-making.
Firstly, GenAI encourages cognitive offloading. According to Gerlich (2025),
cognitive offloading happens when people rely on external tools to do the mental
work for them. GenAI chatbots, with their fast answers and summarised content,
make it easy for users to skip thinking through a problem or evaluating information
deeply. As users get used to the convenience of asking a chatbot rather than
reasoning through a topic, they lose the habit of analysing, comparing, and
questioning different sources of knowledge. Over time, this can lead to shallower
learning and a lack of internalised knowledge, especially among younger users
(Gerlich, 2025).
Secondly, Sarkar et al. (2025) show that GenAI changes the way users engage with
content. In their study, knowledge workers admitted that GenAI made their tasks feel
easier, but it also lowered the effort they put into analysing and evaluating the
information. While these tools can support task completion, they also reduce the
amount of mental effort needed, especially in low-risk tasks. When people trust the
AI’s output without checking its accuracy or thinking through its suggestions, their
ability to problem-solve and think independently is at risk. Sarkar et al. (2025) also
found that over time, users may lose the motivation to challenge or improve on AI-
generated outputs, especially when time pressure and job demands make speed
more valuable than accuracy.
Thirdly, Fügener et al. (2021) highlight the issue of poor delegation between humans
and AI due to a lack of metaknowledge, people often do not know when they should
rely on AI or when to think for themselves. In their experiments, humans performed
worse when asked to delegate tasks to AI, not because they disliked AI, but because
they could not judge their own abilities. This overconfidence in AI and
Varsity Cube 2025 +27 81 278 3372
