×
How to Use AI
May 17, 2026
There are many ideas I leave unfinished in my short essays. I like to use the thoughts page to just jot down ideas that I may have. This creates an environment where I can think outside the box and try to focus on the extremes and crazy ideas without thinking about the details too much. Informal, fluid, and unorthodox.
What does it mean to be a specialist in AI? Everyone keeps saying "learn AI, it's the future", "You need to know how to use AI to have a job in the future" but what exactly does that mean? In this short essay I will talk about what it means to me to leverage AI to become a more productive human.
The STEM Advantage
My background is in Mechanical Engineering. Obviously this is deeply rooted in STEM and I have noticed with the rise in AI I have a critical advantage over others in how I approach AI. I think this advantage is even greater if you are specifically in Computer Science, although you'd be astonished at how many people I talk to that are studying Computer Science that are not bullish on their AI abilities.
The STEM advantage doesn't necessarily come from having a deep knowledge of mathematics or science but from a way that those things make you think about problem solving. I say this because I have seen other people have the same thought process without being in STEM. It's really the clear thinking process of how do I get from A to B as efficiently as possible. How do I manage all of the details and focus what is really the desired outcome. When studying engineering this is all we do. Dry equations and processes make the world look like it's in a state of constant inefficiency because we are looking at the theoretical values of what the ideal condition could be. For me this constantly looks like interpreting data and visual representations of data to make clear, informed decisions. This plays into how I look at using AI because this is what it is inherently very strong at. It makes the barrier of entry to interpreting very large amounts of data into something where you could possibly make a better informed decision. As long as you keep in mind what your problem is and have a defined point where it is solved, AI is an extremely powerful tool.
Using AI
I am going to talk about how I would approach a problem with the former paragraphs information in mind and it will directly show how the STEM advantage and the use of AI combine to make a better problem solver. Then talk about the future of problem solving.
Let's choose an arbitrary problem like increasing the efficiency of a vacuum cleaner. I would first look at what the properties and objectives of a vacuum cleaner are. A vacuum cleaner creates a vacuum that moves mass from an undesirable location to a desirable location, okay this can be represented as some mathematical formula from mass flow rate, power consumption and volume. What are the objectives? To create a vacuum that will take mass (dust and dirt) from one place that is not desired to another that is. Taking this information I could use some equations to describe these relationships and find an ideal state of a vacuum as the main objective. Maybe give a percentage buffer on what I am willing to accept away from that theoretical objective. Possibly scrape data from tests done on high performance or cost effective vacuums that have already been made. Then I could take all of this data and visually represent it through graphs and charts so I can make a better informed decision. We now have our base on our new product and can continue the problem solving process until we have our desired outcome.
The key here is breaking down the problem into its main parts, having clearly defined goals and having a transparent beginning and end goal then using AI to minimize the friction between all of the steps in between. AI could be used at pretty much every step besides the start and the finish.
When looking at it through this lens it appears that it isn't much different as other technological waves. We are still solving a problem and humans still need to know the first question to ask and when the desired outcome is completed. I wasn't thinking about looking up data on the internet on how to create a vacuum 100 years ago but 10 years ago you would be dumb not to do this because it would decrease the amount of time you can complete the project. I believe it's the same thing but with AI this time.
In the future by far the most important realization that will come from using AI and getting objectives done faster is that there is a quality to idea creation. It's not only about knowing how to get from A to B but what exactly are A and B and are they going to benefit my cause the greatest. I think this thought process will dominate everything else in problem solving in the next ten years and beyond. To become good at this process I think you need to understand resource allocation and human desire at a very deep level because this is what has been pushing the economic engine for the last 250 years. It's important now but this will be far more important than traditional problem solving in the future.
What Not to Do When Using AI
I recently had a buddy asking AI for relationship advice. It may have some good answers that might make you feel better but this is way too abstract of a concept for current AI capabilities. Plus, there is no clearly defined desired outcome.
All jokes aside, using AI as the crutch for your critical thinking is a very bad idea since it will hallucinate (give wrong answers). If you have no defined desired outcome then you are really just shooting in the dark which isn't a very great plan when dealing with your actual productivity and the well being of others.
This is why I don't think statements like "humans will be useless in the future" or "I'm going to graduate from college and be a prompt engineer" are valid statements. AI brings people that have worked hard on critical thinking and specializations more fruits of their labor. There is still no way to short cut the critical thinking.
Things to Think About
When using AI, especially one that doesn't have memory of past chats, it's important to think about what the perspective of the AI is. What are the inputs that it is seeing? If it were a person would it understand what the objective is? I often find myself giving short prompts and getting a non-desirable outcome because I'm being lazy and not clearly defining what I want. I think most people would say that the AI is just not very good yet but in my experience the AI is right a lot of the time and I am missing critical pieces of information. So stepping back from the problem and going back to the basics of problem solving can really help here. Making sure that the AI is interpreting your problem in the same way you are is very important. Having a dialogue and asking it if it needs clarifying questions. For me this has resulted in better productivity and reaching my desired outcomes faster.
It's also worth talking about what kinds of inputs are more robust than others. AI doesn't interpret information in the same way as us. It thinks in files and symbols. Learning which ones of these produce the least amount of friction towards your desired outcome will vastly improve your experience in using AI. As a small example of this, I could use the English language to describe what I need done in QuickBooks and this will take a relatively long time and there will likely be some kind of issue because it's not understanding me correctly. But if I upload a .csv file along with a little bit of English there is a greater chance it will give me a better answer faster that is more aligned with my desired outcome. Practicing with AI to find what inputs are going to make you reach your desired outcomes faster is critical.
Conclusion
I talk to a lot of people and there is a general worry that AI is bad. I'm not saying it is not, but I think there is a far greater chance that this new technology is much like every other technology that has come before it. It comes, the general population is scared because it is foreign (companies also make it sound scary for marketing), there are wild assumptions about it, efficient problem solvers use it to get more done and make more profits, and it slowly diffuses into the world because it is a superior way of making more wealth for people and the world. This process has been happening for thousands of years and in my personal position I think this is just the next step in that evolution.