Artificial Intelligence (AI) is now woven into everyday life. From map applications that predict traffic jams, film recommendations on streaming platforms, to chatbots that answer routine questions, AI has become a constant presence. Andrew Ng famously described AI as “the new electricity”, a force that will transform every aspect of society. Just as electricity powered the industrial revolution, AI is driving today’s information revolution.
Yet the rise of AI does not mean humans can stop thinking. On the contrary, the more decisions are assisted by machines, the more vital our ability to evaluate, question, and filter information becomes. This is where critical thinking plays its role.
Critical thinking is not confined to academia; it is a life skill. It enables us to distinguish fact from opinion, assess arguments, and make wiser choices. In the context of AI, critical thinking acts as a safeguard: machines provide answers, but humans must judge whether those answers are accurate, relevant, and useful.
Alongside this, a new concept has emerged: prompt engineering. This is the craft of designing questions so that AI produces responses tailored to our needs. Prompt engineering is, in essence, an exercise in critical thinking. It requires clarity, precision, and awareness of context. In other words, AI encourages us to be more deliberate in how we ask, not merely in how we receive.
AI as a New Literacy
We often take reading and writing for granted. Yet centuries ago, literacy was a privilege enjoyed by only a few. Today, Andrew Ng argues that coding—especially coding for AI—is becoming a new form of literacy. Just as words allow humans to communicate deeply with one another, code allows us to communicate with machines. And as machines become more central to our lives, this kind of literacy grows in importance.
AI literacy is not about building video games or websites, but about using data to solve real problems. Imagine a pizza shop owner who wants to predict how many Hawaiian pizzas to prepare each Sunday. With a simple regression model, he could forecast demand, reduce waiting times, and manage his supplies more efficiently. This is a small but powerful example of how AI literacy can improve everyday life.
The key difference between traditional coding and AI-oriented coding lies in the approach. Traditional coding tells a computer exactly what steps to follow. AI coding, on the other hand, teaches the computer to learn patterns from data. This shift makes AI literacy even more valuable, because it enables people in diverse professions—from healthcare to agriculture—to harness insights from data without needing to become expert programmers.
Here is where critical thinking comes in. Learning to code for AI is not enough; we must also learn to question the data, the models, and the outcomes. Is the dataset representative? Are the predictions reliable? Could there be hidden biases? Without critical thinking, AI literacy risks becoming mechanical, producing results that may look convincing but lack substance.
In short, AI literacy is more than a technical skill. It is a mindset that combines coding with questioning, data with judgement, and automation with human responsibility. Just as literacy in language enriched society, literacy in AI—guided by critical thinking—can enrich our digital age.
Prompt engineering is the art of crafting questions or instructions so that an AI system, such as ChatGPT, produces answers that are relevant, clear, and suited to the user’s needs. In essence, the machine responds to whatever we ask of it. If the question is vague, the answer may drift off course. By applying prompt engineering, we learn to design precise queries, which in turn leads to higher‑quality results.
This practice often involves specific approaches. For instance, one might give highly detailed instructions such as “Write a summary of this news article in one paragraph”. Another method is to assign the AI a role, for example “As a history teacher, explain the Second World War”. A further technique is to use a keyword to guide the response, such as “Compose a poem with the keyword ‘love’”. There is also the option of ensuring that the answer remains consistent with the context provided, so that the output does not contradict itself.
The connection between prompt engineering and critical thinking is strong. When we construct a prompt, we are required to frame the question clearly, establish the right context, and then evaluate the answer that the AI delivers. This process trains us not to accept responses at face value, but to consider whether they meet the instructions, whether errors are present, and whether adjustments are needed.
Thus, prompt engineering is not merely a technical trick to “outsmart” a machine. It is a practical exercise in critical thinking. By learning to ask more intelligently, we ensure that the answers generated by AI are more useful and more closely aligned with real‑world needs.
Let’s make prompt engineering feel tangible by showing how it works in everyday life.
In studying, imagine a student revising history. Instead of asking AI “Explain World War II”, which is too broad, they might say: “As a history teacher, summarise the causes of World War II in three clear points for a beginner.” This role‑based prompt guides the AI to produce a structured, accessible answer, making learning more effective.
In professional work, consider someone drafting a report. If they simply ask “Write about climate change”, the output may be unfocused. But with prompt engineering, they could say: “Generate a concise executive summary on climate change impacts for business leaders, highlighting risks to supply chains.” Here, the instructions and context ensure the AI delivers something tailored to a corporate audience.
In leisure, think of a person wanting creative writing. Instead of “Write a poem”, they might say: “Compose a sonnet about summer evenings, using the seed word ‘nostalgia’.” By specifying the form and theme, the AI produces a piece that feels more personal and enjoyable.
In the world of journalism, prompt engineering also plays a crucial role. A journalist cannot simply ask an AI “Write an article about the economy”, because the result may be too general and unfocused. With prompt engineering, the request can be sharpened, for example: “As an economics reporter, produce a concise summary of the impact of inflation on food prices in Indonesia, written in a neutral and informative style.” This kind of instruction ensures the output is closer to the standards of clear, balanced reporting.
Prompt engineering can also assist with information verification. Suppose a report contains conflicting figures about the number of casualties in an incident. A journalist might ask: “Check the consistency of the data in this report and highlight any contradictions.” This is similar to self-consistency prompting, which is useful for detecting inconsistencies in text.
Another application is in interview preparation. A journalist could request: “As a senior editor, generate five critical questions for an interview with an energy expert about the transition to renewable energy.” This helps to frame sharper, more relevant questions that lead to stronger interviews.
Ultimately, in journalism, prompt engineering is not just a technical tool but a practice of critical thinking. It trains journalists to phrase questions clearly, maintain objectivity, and ensure that the information presented is accurate and valuable to the public.
Across these examples, the common thread is critical thinking. Prompt engineering forces us to pause, clarify what we truly want, and express it precisely. The better we think, the better the AI responds — whether we are learning, working, or simply having fun.
Critical Thinking in Learning AI Skills
Learning AI is not just about memorising formulas or mastering coding libraries. It is about knowing what to learn, why it matters, and how to apply it. Andrew Ng reminds us that the field of AI is vast, with more research papers published than anyone could read in a lifetime. This is where critical thinking becomes essential: it helps us prioritise, select, and evaluate the skills that will truly support our goals.
For example, a beginner might be tempted to dive into advanced calculus or complex neural architectures straight away. Yet critical thinking encourages us to ask: Do I really need this knowledge now? Will it help me build the kind of projects I want? In many cases, understanding the basics of linear regression, logistic regression, or decision trees is far more useful at the start than chasing the latest research trend.
Critical thinking also plays a role in how we approach errors and setbacks. When a model fails to converge or produces strange results, the easy reaction is frustration. But a critical thinker will pause and ask: Is the dataset flawed? Are the assumptions valid? Could bias or noise be affecting the outcome? This habit of questioning transforms mistakes into opportunities for deeper learning.
Moreover, critical thinking helps us resist the temptation to learn passively. Reading random web pages or tutorials may feel productive, but without a structured approach, the knowledge gained is often fragmented. A critical learner will choose coherent courses, evaluate the quality of resources, and gradually move on to research papers once the foundations are solid.
In short, learning AI is a lifelong journey, and critical thinking is the compass that keeps us on track. It ensures that we do not drown in information overload, but instead build skills that are relevant, applicable, and resilient in a rapidly changing field.
Choosing the Right AI Project
Building AI skills alone is not enough; what truly matters is how we select the right projects to pursue. Andrew Ng highlights that many people fall into the “Ready, Aim, Fire” mindset — spending too long planning before ever taking action. In contrast, the “Ready, Fire, Aim” approach is often more effective: start with a small project, then learn and adjust along the way.
Here, critical thinking becomes essential. Before beginning a project, we must ask ourselves: Is this problem genuinely important? Is the available data sufficient? Will an AI solution create real impact? These questions help us avoid projects that may look technically impressive but fail to address meaningful needs.
Take, for example, a company tempted to build a sophisticated chatbot simply because it is trendy. Critical thinking would prompt the question: does this chatbot truly solve customer problems, or does it risk adding confusion? In this way, AI projects become more than technological experiments; they become solutions that add genuine value.
Critical thinking also helps us assess risks and biases. If a dataset only represents one group, the results may be skewed. A critical thinker will recognise this and seek ways to improve the dataset so that it is fairer and more representative.
In short, choosing an AI project is not just about creativity; it is about careful judgement. With critical thinking, we can select projects that are relevant, realistic, and impactful—whether in business, education, or everyday life.
AI as a Partner, Not a Substitute for Critical Thinking
Artificial Intelligence is often portrayed as a tool that can “think” for us. Yet the reality is that AI does not replace human judgement; it complements it. Machines can process vast amounts of data, identify patterns, and generate suggestions at incredible speed. But they cannot decide what truly matters, nor can they weigh ethical consequences or social impact. That responsibility remains firmly in human hands.
Critical thinking ensures that AI becomes a partner rather than a master. For instance, when AI recommends a medical treatment, it is the doctor’s critical judgement that determines whether the recommendation is safe, appropriate, and tailored to the patient. Similarly, when AI suggests a business strategy, it is the manager’s responsibility to evaluate whether the plan aligns with company values and long‑term goals.
This partnership works best when humans and AI play to their strengths. AI excels at computation, prediction, and automation. Humans excel at reasoning, empathy, and ethical reflection. Together, they form a powerful combination: AI provides options, and humans decide which option is right.
The danger lies in over‑reliance. If we accept AI outputs uncritically, we risk amplifying biases, overlooking errors, or making decisions that lack context. By contrast, when we approach AI with a questioning mindset—asking “Is this accurate? Is this fair? Is this useful?”—we transform it into a tool that enhances rather than diminishes our thinking.
In short, AI should be seen as a collaborator. It can accelerate our work, broaden our perspective, and spark creativity. But it is critical thinking that ensures those benefits are channelled responsibly, keeping human judgement at the centre of every decision.
Building a Culture of AI and Critical Thinking
AI is not merely about technology; it is also about culture. For AI to be genuinely useful, schools, workplaces, and communities must foster habits of using AI alongside critical thinking. Without such a culture, AI risks becoming a superficial tool, or worse, a source of bias and error.
In schools, this culture can be nurtured by encouraging students not simply to accept answers from machines, but to question and evaluate them. For instance, when AI provides a summary of a historical event, teachers can guide students to compare it with other sources and discuss whether the summary is complete and accurate. In this way, AI becomes a means of practising critical thinking rather than replacing the teacher.
In the workplace, building this culture means encouraging employees to treat AI as an assistant, not a sole decision‑maker. A business analyst, for example, might use AI to generate market forecasts, but must still judge whether those forecasts are realistic, data‑driven, and aligned with company strategy. This critical culture ensures that business decisions remain grounded in human judgement.
Within communities, the culture of AI and critical thinking can be expressed through digital literacy. People need to understand that not all AI outputs are correct or neutral. By developing the habit of asking “Does this make sense? Is there supporting evidence elsewhere?”, communities become more resilient against misinformation or bias that may arise from AI systems.
In short, building a culture of AI and critical thinking means positioning AI as a partner that strengthens human capability rather than weakens it. Such a culture ensures that technology is used wisely, fairly, and for the benefit of all.
The Future of AI and Critical Thinking
The future of Artificial Intelligence will not be defined solely by faster algorithms or larger datasets. It will be shaped by how humans integrate critical thinking into their use of AI. As technology advances, the challenge is not whether machines can do more, but whether people can continue to question, evaluate, and guide those capabilities responsibly.
In education, AI will likely become a common tool for personalised learning. Yet the real value will come when students are taught to interrogate AI’s answers, compare them with other sources, and reflect on their accuracy. This ensures that AI nurtures independent thought rather than passive consumption.
In the workplace, AI will automate routine tasks and provide sophisticated insights. But critical thinking will remain the safeguard that prevents blind reliance. Managers and professionals will need to ask: Does this recommendation align with our values? Is the data unbiased? What are the long‑term consequences? Without such reflection, efficiency could come at the cost of fairness or sustainability.
In society at large, AI will influence politics, media, and everyday decision‑making. The danger is that misinformation or bias could spread more quickly than ever. A culture of critical thinking — questioning sources, demanding transparency, and evaluating context — will be essential to protect democratic values and social trust.
Ultimately, the future of AI is inseparable from the future of human judgement. Machines may grow more powerful, but it is critical thinking that ensures they serve humanity wisely. The partnership between AI and human reasoning will define whether technology becomes a force for progress or a source of division.
Conclusion: Towards a Critical AI Literacy
Our journey through AI literacy, prompt engineering, critical thinking, and project selection reveals a single thread: AI will only be truly valuable if it is used with awareness and sound judgement. AI literacy teaches us to grasp the fundamentals, prompt engineering trains us to communicate effectively with machines, while critical thinking ensures we do not simply accept answers but also evaluate, question, and refine them.
In education, AI can accelerate understanding, yet teachers and students must still test its outputs. In workplaces, AI can speed up analysis and automation, but final decisions must consider values, strategy, and long‑term impact. In society, AI can broaden access to information, but critical literacy is needed to avoid bias or misinformation.
Ultimately, critical AI literacy is not just a technical skill but a mindset. It requires us to continually ask: Is this accurate? Is this fair? Is this useful? With such an attitude, AI becomes a partner that strengthens human capability rather than replacing it.
The future of AI will continue to evolve, but the future of humanity depends on our ability to keep critical thinking at the centre of every interaction with technology. In doing so, we become not merely users of AI, but guides who ensure that this technology delivers genuine benefit to life and society.
References
- Ng, Andrew. Machine Learning Yearning. DeepLearning.AI, 2018.
- Russell, Stuart, and Peter Norvig. Artificial Intelligence: A Modern Approach. 4th ed., Pearson, 2020.
- Floridi, Luciano. The Ethics of Artificial Intelligence. Oxford University Press, 2021.
- Dignum, Virginia. Responsible Artificial Intelligence: Designing AI for Human Values. Springer, 2019.
- Mitchell, Melanie. Artificial Intelligence: A Guide for Thinking Humans. Penguin, 2019.
- Klein, Gary. Sources of Power: How People Make Decisions. MIT Press, 1998.
- Facione, Peter. Critical Thinking: What It Is and Why It Counts. Insight Assessment, 2015.
- Silver, Nate. The Signal and the Noise: Why So Many Predictions Fail – But Some Don’t. Penguin, 2012.
