I'm a computer science student, and I've become stuck on a basic question: how do people come up with project ideas? For the past two years, I've usually asked AI for suggestions or started with an existing popular project. I don't blindly accept the first idea—I evaluate its usefulness, impact, and value for my portfolio—but AI or online research is still always my starting point.
I don't want to depend on AI to do all of my thinking. Before tools like this existed, how did developers and students discover projects worth building? How can I tell whether an idea is useful enough for a portfolio without dismissing interesting ideas too early? I'm looking for a practical way to find, evaluate, and develop project ideas so I can become more independent and better prepared for future work or research.
5 Answers
Look for problems in your own life and in the lives of people around you. Notice repetitive tasks, manual work, confusing processes, or existing tools that almost solve a problem but could be better. You can also speak with classmates, professors, or small-business owners and ask what tasks waste their time. If several people mention the same annoyance, build a small solution and let actual use suggest the next features.
Start by separating the purpose of the project from its portfolio value. A project can be for learning, experimentation, research, or demonstrating a particular skill; it does not have to accomplish all of those at once. Ask what you actually need to build, who would use it, how you would demonstrate it, and what part of the process you would be proud to explain. You can decide later which experiments deserve a place in your portfolio.
Most of my projects are meant to show that I’ve done something and eventually support my portfolio. The difficult part is that I either choose something unrelated or lose confidence as soon as I try to work on a relevant idea. I often like the subject but still end up feeling stuck and unmotivated.
Professors and people working in fields you’re interested in can be useful sources of direction because they know which problems and skills currently matter. You can also choose an existing application and make a focused improvement or a simpler version. A small web tool, local database, data pipeline tracker, or automation project can be worthwhile when it solves a real problem and gives you something concrete to discuss.
The first step may be to stop using AI for idea generation for a while. If you’ve relied on it to do that part of the thinking, coming up with ideas yourself may feel difficult at first because it’s a skill you haven’t practiced recently. Keep a notebook or digital list and write down every annoyance, curiosity, or improvement you notice. Many ideas will be bad, but the useful ones will become easier to recognize with practice.
That makes sense. I didn’t realize how much I had trained myself to outsource that first step. I used to come up with ideas more naturally, so I probably need to rebuild the habit.
Ideas often come from existing software, coursework, hobbies, and everyday frustrations. Whenever you think, “This would be easier if a program did X,” record it immediately. Keep an idea list and review it occasionally instead of expecting one perfect concept to appear on demand. Most ideas will go nowhere, but a few will become interesting once you explore them. Start with a small version, then let the experience of using it reveal what should be improved next.
It doesn’t need to be completely original. Improving an existing tool, automating a boring task, or building a small internal dashboard can still be a strong project if you can explain the problem, your design choices, and what you learned.