Is AI Making You Dumb? 5 Habits to Take Back Control

0:00You are being manipulated, like a marionette, not by a person, by artificial intelligence. Every time AI thinks for you, writes for you, researches for you, or makes decisions for you, it's like another string is being pulled. And the scariest part? Most of us don't even realize it's happening. We think we're becoming more productive. But what if we're actually becoming less independent,

0:23less creative, and less capable of thinking for ourselves? And here's something I don't like admitting. I think it's happening to me too. Think about it. When was the last time you wrote something important without using ChatGPT, Copilot, Claude? When was the last time you read an entire research paper instead of asking AI to summarize it? When was the last time you solved a simple math problem without using a calculator or asking AI? Or spent hours researching a topic, comparing

0:52sources, questioning information, and developing your own conclusions? Today we just ask AI, and within seconds we have an answer, an essay, a business plan, a financial analysis, a recommendation, even an entire research paper. And the scary part? Sometimes we don't even know if the answer is correct. We just trust it, because it sounds professional.

1:15Now, don't get me wrong. I love artificial intelligence. I use it almost every single day. In fact, I've built this YouTube channel around helping people use AI in real situations. But I've been thinking about something that genuinely worries me. What happens to our ability to think when technology does most of our thinking for us?

1:36And what will happen to the next generation? A generation that might have access to more information than any generation in history, but potentially fewer opportunities to develop the skills needed to question and understand that information? Let me explain what I mean. I completed my Master of Global Management in 2024. I also earned my Project Management Professional Certification, the PMP,

2:00in 2023. And I've been asking myself a controversial question. Is the education I completed before generative AI became so powerful worth more in terms of what I actually learned than completing the same education today? Now, don't misunderstand me. I'm not saying a Master's degree today is worth less. And I'm definitely not saying people earning their certifications now aren't knowledgeable. There are incredibly intelligent people studying

2:28today. But I do think the learning experience has changed dramatically. When I was doing academic work, I had to spend time researching, reading articles, organizing information, comparing arguments, and writing. Of course, tools existed, and ChatGPT was already available, although from what I saw, only a handful of people were using it during part of my studies. The most I used was something called Grammarly,

2:57and I used it mainly to catch mistakes in my English. English isn't my first language, so Grammarly was genuinely useful for me. It would identify grammatical mistakes, suggest better ways of structuring sentences, and help me understand what I was doing wrong. And I actually think it improved my English, because I was still doing the writing, I was still developing the arguments, I was still reading

3:22the academic papers, and organizing my thoughts. The tool was helping me improve something I had already created. But today? Today, someone can approach the same assignment in a completely different way. Imagine someone has to write a 3,000 -word academic paper. They can open ChatGPT, describe the assignment, ask for an outline, generate sections, improve the writing,

3:46and even get suggestions for academic references. And then, if that isn't enough, they can take the entire paper and paste it into another AI tool, a so-called AI humanizer. And what does that tool promise to do? Remove the AI sound from the paper, make it sound more natural, make it look as if a human wrote it.

4:07Think about how strange that is. We're using artificial intelligence to write something, and then using another artificial intelligence tool to make it look like artificial intelligence wasn't involved. Strange, right? Yes. But this is exactly the type of workflow some of these tools are promoting. And it raises a question about what we are actually measuring when we evaluate someone's work. A task that could have required days of reading and thinking might now

4:36produce a polished draft in minutes. But here's my question. Did that person actually learn the subject? I actually spoke about a related issue in a video a few months ago, where I discussed how difficult managing people can become in the AI era. I'll link that video in the description. One of the challenges I discussed is what happens when an employee delivers something impressive using AI, but then struggles to explain

5:01the results. Imagine you're a manager. Someone on your team delivers a beautiful report. It has charts, recommendations, financial analysis, and everything looks fantastic. But when you ask a simple question like, can you explain how you reached this conclusion? They can't, because AI generated the analysis. Now, I'm not saying everyone using AI is doing this, but I do think managers increasingly need to evaluate whether employees actually

5:30understand the work they're submitting. And the same applies to education. Can someone explain their argument without looking at the screen? Can they defend their conclusions when challenged? Can they tell when an AI generated reference is fake or recognize that a paper sounds excellent but doesn't actually make sense? That's what worries me. And the same thing applies to professional certifications. When I studied for my PMP, I had to learn project management concepts, understand scenarios,

5:57and practice applying that knowledge. Today, someone can have AI explain almost any project management concept instantly. And depending on their experience and preparation, someone might pass a certification exam while having very little real -world experience applying those concepts. Now, the PMP itself has professional experience requirements. I'm not suggesting someone can legitimately earn a PMP without meeting those requirements. But my bigger point is that passing an exam, understanding a textbook, and actually being able to do the job are three different things. Imagine you know all the project management

6:36terminology. You know what a risk register is. You know what a stakeholder management plan is. You know what earned value management means. But you've never had to deal with an unhappy executive, a contractor missing deadlines, or a project that's running out of money. And then you become responsible for a real project. You can ask Copilot to create a recovery plan.

6:59It might produce a beautiful document. But can you determine whether the recommendation actually makes sense? Can you defend it? Can you recognize what AI doesn't know? I don't necessarily think older qualifications should automatically be worth more on paper. But I do think the ability to learn independently, apply knowledge, and make difficult decisions is becoming more valuable.

7:23And that's a very important difference. Here's another concern. I think we're slowly losing the habit of doing real research. Not everyone, of course. But think about how many of us find information now. Before, if you wanted to learn something, you might read a book, search through several websites, find academic papers, and compare different perspectives. Sometimes the research was frustrating.

7:46You couldn't find the answer. One article contradicted another. You had to decide which sources were reliable. But that was part of the learning process. You weren't just collecting information. You were developing judgment. Now we can type, research this topic, and give me the five most important conclusions. And there it is. Five beautiful bullet points.

8:09Maybe a table, maybe even some citations. But did we read the original sources? Did we verify the statistics? Did we understand the limitations? Or did we just copy the answer into a presentation? I actually made another video showing a different way of doing research on the internet using AI. And I think it's one of the most important concepts I've shared on this

8:31channel. The funny thing is, that video didn't get many views. But I really believe the concept deserves more attention. In that video, I show how we can use AI as part of a research process while also looking for feedback from real people, not just relying on what the AI tells us, but comparing its findings with actual experiences, opinions, discussions,

8:57and different perspectives available online. Because imagine you are researching whether a product is reliable, AI might summarize the product's features and specifications. But what are people who have used that product for two years saying? What problems have they experienced? Are there recurring complaints? Are there disagreements? And which opinions are actually based on first

9:19-hand experience? That information can change your understanding. Of course, feedback from real people isn't automatically reliable either. People can exaggerate, misunderstand things, or have completely different experiences. That's why you need to compare sources instead of accepting one opinion as the truth. Now, this process takes longer. It's not as easy as opening ChatGPT and asking what's the best option. You actually have to investigate, compare information, question different perspectives, and sometimes read things you don't agree with. But I believe the results can be very different and

9:55potentially much more useful. I'll also link that video in the description, because I genuinely think it's worth watching, even if it wasn't one of my most popular videos. And reading is another example. We have tools that can summarize an entire book, a hundred-page report, or hours of meeting transcripts in seconds. That's genuinely useful. I use these features myself. But when summarizing becomes the only way we consume information, we

10:24miss the details, the arguments, and the deeper context. A summary tells you what someone concluded. Reading the full material can help you understand how they arrived at that conclusion. Those are two very different things. And sometimes the most important insight isn't in the summary, it's in the part you would have questioned while reading.

10:46Now, let's talk about something even more serious. Government. I live in Canada, and artificial intelligence is already becoming part of how governments work. The Canadian federal government has made Microsoft Copilot available under certain conditions, and departments have been adopting AI tools to support their employees. Here's something I found particularly interesting. The government of Canada actually confirmed that Microsoft

11:11Copilot was used during the development of its own federal AI strategy. AI helped summarize research, organize feedback, and draft documents. And according to the government, those outputs were reviewed by human analysts. That's an important example of how these tools can be used responsibly. But it also shows how deeply AI is entering government operations. Now, let's look at the United States. In 2025, the US Government Accountability Office reported that generative AI use cases across 11 selected federal agencies increased from 32 in 2023 to 282 in

11:532024. That's nearly nine times as many. And we're not just talking about helping someone write an email anymore. In June 2026, the US military announced an initiative called Agent Network. The idea is to use AI-enabled systems to analyze intelligence and present options to military commanders much faster. We're talking about information that could influence battlefield decisions. And in March 2026, Reuters reported that the Pentagon was making Palantir's Maven AI system a core part of its military capabilities. Maven helps analyze intelligence and identify potential

12:34threats and targets. These are serious applications. Now, I want to be very clear. These examples do not prove that military leaders are blindly following AI recommendations. Human oversight remains an essential part of these systems. But they show us something important. We have reached a point where AI is helping inform decisions that can have enormous consequences. So my question is, what happens if the people responsible for those decisions gradually become less capable of evaluating the

13:09recommendations themselves? Imagine a leadership team is considering a major investment. Someone asks AI to analyze the situation and recommend the best option. AI generates a professional -looking report with a strong recommendation. The language is confident. The formatting is excellent. The arguments sound reasonable. So everyone agrees. But what if the analysis is based on incomplete

13:34data? What if the AI misunderstood one of the assumptions? What if it overlooked something important that an experienced employee would have recognized? And what if nobody in that meeting has enough knowledge to challenge the recommendation? That's the real danger. I'm not saying leaders should avoid AI. I think they absolutely should learn how to use it. But AI should support decisions, And a beautiful presentation isn't

14:01evidence of good analysis. I believe leadership in the AI era will require something extremely important, the courage and knowledge to challenge the machine. Now, here's a smaller example that I think says a lot, basic math. How often do we calculate something in our heads anymore? A percentage, a discount, a simple multiplication. Sometimes we use a calculator for something we could probably solve

14:26ourselves. Now we also have AI. We don't even need to figure out which formula to use. We just describe the problem, and AI gives us the answer. Again, that's incredibly convenient. But what happens when we stop practicing? It's like going to the gym. If you never challenge your muscles, you shouldn't be surprised when they become weaker.

14:49I think some thinking skills work similarly. The less we practice them, the less confident and capable we may become. And what about writing? Before AI, writing a good email, developing a convincing argument, or preparing a professional report required practice. Today, you can type three sentences and ask ChatGPT to turn them into a polished document. I do this myself, and sometimes I stop and think,

15:15would I still be able to write this as well without AI? That's an uncomfortable question, because I don't want to become someone who needs artificial intelligence to communicate a simple idea. I want AI to improve my abilities, not slowly replace the abilities I already have. And then there's creativity. This one really makes me think. We can now generate images, videos, music,

15:41presentations, logos, websites, and entire stories with a simple prompt. That's amazing. Someone who couldn't create these things before now has the ability to bring an idea to life. But there's another side. When everyone asks AI for ideas, everyone can start receiving similar suggestions, the same presentation structures, the same LinkedIn posts, the same professional emails,

16:05the same YouTube video hooks, even the same phrases. Have you noticed how much AI-generated content sounds alike? What happens when we stop sitting with a problem long enough to come up with something original? Creativity often comes from curiosity, experimentation, frustration, and mistakes. Sometimes your first idea is terrible. Your second idea isn't much better. But by the fifth attempt, you've created something genuinely interesting. If we always accept the first AI-generated suggestion,

16:35we might skip that entire creative process. Now, I don't believe creativity is over. AI can actually expand creativity when we use it to experiment and explore. But I worry about a future where people become excellent at generating content and less experienced at creating ideas of their own. Just because you can generate something doesn't necessarily mean you

16:56created something meaningful. Now, I want to make something very clear. I'm not making this video because I think I'm smarter than people using AI. I'm one of them. I use ChatGPT, Copilot, and other AI tools constantly. I use them to organize my thoughts, improve my writing, research topics, analyze information, and create content. And sometimes, I'm probably relying on them too much. Sometimes, I'll ask AI a question before I've even tried to answer it myself. Sometimes, I'll ask it to rewrite something I could have written. And sometimes, I catch myself accepting an answer because it

17:33looks reasonable, without spending enough time validating it. So, I'm fighting this habit too. I'm trying to challenge myself, to read more, to think before asking, to understand why an answer is correct, to do some things manually, even when AI could finish them faster. Because I don't want to become more productive while becoming less capable. And I don't want to teach people how to use AI, without also

18:01encouraging them to think critically about what AI produces. That's especially important to me, because this channel is about learning AI by doing real work, not just copying prompts, not just generating impressive results, actually understanding what we're doing. So, what can we do? I don't think the solution is to stop using AI. That's unrealistic, and I don't think it's necessary.

18:26Instead, here are five habits I'm trying to practice myself. Number one, think before you prompt. Before asking AI to solve something, take 30 seconds and try to develop your own answer, even if your answer isn't perfect. Then ask AI to compare its response with your reasoning. That way, you're practicing thinking instead of immediately outsourcing

18:49it. Number two, ask AI to challenge you. Don't always ask ChatGPT to agree with your ideas. Try saying challenge my assumptions. Tell me where my reasoning is weak. Give me the strongest arguments against my conclusion. Use AI as a critical thinking partner, not as a machine that constantly tells you you're right. Number three, verify the information.

19:13This one is extremely important. If AI gives you a statistic, check the source. If it summarizes a research paper, open the paper. If it produces a financial analysis, verify the numbers. And if it recommends an important business or government decision, examine the assumptions before accepting the recommendation. Remember, AI can produce incorrect information with a great deal of confidence. You need to know enough to recognize when something doesn't make sense. Number four,

19:44learn to explain things without AI. After using AI to learn something, close the chat. Try explaining what you learned in your own words. If you can't explain it, go back and study it again. For example, if Copilot helped you build an Excel formula, try recreating it yourself. If ChatGPT helped you understand a project management concept, explain

20:07that concept without reading the answer. The goal is to make the knowledge yours, not just something stored in a conversation. And number five, use AI as a coach, not a replacement. This is probably my favorite. Instead of asking AI to give you the solution, ask it to teach you. You can say, act as my learning coach.

20:27Give me a problem. Let me try solving it and then evaluate my reasoning. Don't give me the answer immediately. Imagine doing that for Excel, SQL, project management, communication, or even learning another language. Now you're using AI to build your skills instead of avoiding the work required to develop them. And I think that's how we should be using these tools more often.

20:50So what does the future look like? I think we're entering a world where generating information will become incredibly easy. Everyone will be able to produce professional reports. Everyone will be able to create presentations. Everyone will have access to tools that can write code, analyze data, and suggest solutions. But here's the question. If everyone can generate an answer, what makes someone valuable? I believe it's the ability to recognize whether that

21:19answer is correct, the ability to ask the right questions. The ability to understand the context, the ability to recognize risks that AI might overlook, and the ability to make a decision and take responsibility for it. You've probably heard people say, AI won't replace people. People who use AI will replace people who don't. I think that statement is incomplete, because simply knowing how to use

21:46ChatGPT isn't going to be enough. Almost everyone will eventually know how to ask AI a question. The real difference will be between people who accept whatever AI produces, and people who can evaluate, challenge, improve, and apply its recommendations. AI might not replace you. But someone who understands your work, knows how to use AI, and can

22:10validate its sensors may have a real advantage over you. And that applies to me as much as anybody else. That's why I believe we have to keep learning. Not just learning new AI tools. Learning the subjects we're actually using those tools to work on. Learning how to research, how to communicate, how to question information, how to solve problems, and how to think independently. Because the future doesn't need millions of people who

22:37can copy and paste an AI-generated answer. It needs people who understand what those answers mean. So is AI making us dumb? I don't think we have enough evidence to say AI is making everyone less intelligent. But I do believe we're at risk of losing important habits if we stop practicing them. And I see that risk in myself. I don't want to look back in 10 years and realize that I've become

23:02excellent at using technology, but worse at thinking independently. I want to become better because of AI, not dependent on it. And that's the challenge I want to leave you with. Next time you open ChatGPT or Copilot, ask yourself one question. Am I using AI to learn something or am I using AI to avoid thinking?

23:23Because there's a huge difference. And remember, the future won't belong to the people who get the fastest AI answers. It will belong to the people who know which answers to trust. Don't outsource your brain. Train it. Now I'm curious to hear what you think. Do you believe your education or professional experience from before AI

23:44has a different value today? And do you think AI is making us smarter or are we becoming too dependent on it? Let me know in the comments. I genuinely want to hear different opinions. If you enjoyed this video, subscribe to TERUIO. Here we learn AI by doing real work. See you in the next one.