We are entering another level of technology development: generative artificial intelligence (AI) and automation. All these tools help us save time, be more productive and precise in our output. But does it really help?

To be more productive, generative AI such as ChatGPT and Perplexity, for instance, will surely be the 24-hour assistant to humans with constant results. However, we still give prompts to deploy the criteria. It can be prompted by written or verbally deployed, that known as vibe coding. Thus, AI can’t stand alone and proceed without a request or code given.

Reflecting on this, our brain is still in charge of providing and developing ideation. And apparently, this convenience spoils most of its users. Sometimes, users generate AI by not giving detailed prompts as a reflection of reliance on AI, resulting in repetitive style results. Here, these AI results can’t be as natural and authentic as expected.

AI changes how humans think and utilise these automation hypes. Does it help, or is it just hype for short? Let’s find out more.

How Generative AI is Changing Our Lives

Initially, the purpose of generative AI is to simplify complicated tasks and help humans provide seamless operations to be productive. However, the results need to be redefined, as not all AI results will be identical as expected. 

Let’s talk about the impact of dependence on generative AI. Apparently, this makes humans no longer think critically and authentically. Everything is automatically standardised to achieve the final work results that match the output and are perfect without defects.

This is proven by the results of a survey conducted by Hao Pink (Hank) Lee, a researcher who collaborated with six researchers from Microsoft, on 319 workers who collected 936 tasks in a professional worker scenario working relying on AI, in the survey “The Impact of Generative AI on Critical Thinking: Self-Reported Reductions in Cognitive Effort and Confidence Effects From a Survey of Knowledge Workers”.

Participants were asked to assess their critical thinking performance when working with GenAI. They went through three stages: using AI to create content, using AI to find or work with information, and using AI to get advice. About 60% (555 of 936) of sample tasks handled by participants self-reported that they triggered critical thinking activities based on task confidence and when reflecting on the work done using generative AI tools.

Participants activate critical thinking activities when working on the AI tools canvas, which AI then generates. From here, we can see that our critical thinking activities are used as AI materials to process information, combined with information data that is already embedded in AI.

Critical Thinking and Cognitive Skills

Generative AI assistance provides convenience or serves as a shortcut to human critical thinking. When receiving a prompt from a human, AI shifts the critical thinking process into three stages:

• For Knowledge and Comprehension

Shifting effort from information gathering to information verification.

• For Application

Changing critical thinking from problem-solving to AI response.

• For Analysis, Synthesis, and Evaluation

Redirecting focus from task execution to task stewardship.

Humans tend to execute light tasks as well as supervisory activities, such as revising AI results. Thus, the human role disappears at the stage of creating work or content, but at the stage of providing a brief or prompt and executing the results.

Doing this is considered less risky or critical engagement for the outputs. Humans rely on AI because they feel the results provided are valid without spending more time than usual.

Positive and Negative Impact

Relying on generative AI does not provide complete relief. The positive impact clearly helps as a shortcut in operations, but it dulls human creativity and critical thinking in creating work. Recognise the positive and negative effects, still according to the survey conducted by researcher with Microsoft researchers:

• Effort Reduction
  • Automated tasks (e.g., drafting text or generating documents) saw reduced effort for participants in synthesis activities (129/319 examples).
  • 48/319 participants reported reduced effort in analytical tasks due to GenAI.
  • 129/319 participants experienced reduced effort in synthesis tasks like creating drafts or responding to emails.
• Effort Growth
  • Tasks requiring analytical skills (e.g., translating intentions into specific queries or revising prompts) increased users’ perceived effort, as noted by 45/319 participants.
  • 45/319 participants noted increased effort in creating effective queries or prompts for GenAI tools like ChatGPT.
  • 48/319 participants found the need for steering and revising AI responses to add effort.

Overreliance on GenAI

Does over-reliance on generative AI have a significant impact on human work outcomes? Humans experience self-doubt when working on tasks independently. Especially when it comes to technical matters such as drafting legal documents.

AI can be helpful, but it is not recommended to rely entirely on it for the best results at this time. Overreliance on generative AI is documented as a risk, especially in high-stakes work environments. When errors occur in generating AI, this can lead to negative outcomes and misinterpretations.

Conclusion

Considering AI as your full assistance is not wrong, as the industry will adapt to this. It will probably be adopted into the education sector as well in the future. However, overuse and overreliance on AI can reduce human critical thinking engagement. Therefore, human intervention still needs to be involved and take control of quality and relevancy guaranteed.

---

Source:
Lee, H., Sarkar, A., Tankelevitch, L., Drosos, I., Rintel, S., Banks, R., & Wilson, N. (2025). The Impact of Generative AI on Critical Thinking: Self-Reported Reductions in Cognitive Effort and Confidence Effects From a Survey of Knowledge Workers. Microsoft Research.