Usually when people have a question about something the first place they go is to the internet, and usually I do too. However, I had an eye-opening experience where the internet was not helpful and it exposed my lack of knowledge.

In my senior year of high school, I applied and was accepted as an ASPIRE student (aka unpaid intern) to the Johns Hopkins Applied Physics Lab. My mentor, Philip Graff, was a genius. He was a part of the group that published the paper on gravitational waves in 2016 (a tremendous scientific achievement) and is one of the best in his field of clustering big datasets. Coming into the internship, my coding experience was mediocre. I had learned the basics of java and python, but that was about it. As I began, my mentor introduced me to neural networks, which are advanced algorithms that interpret data by running training and test data and improve themselves by learning off mistakes. I know it sounds advanced, but I was thinking the same way. I jumped from what was an intro to programming class to graduate school level coding.

I started off learning fast. I was starting to write my own mini neural networks with some step by step instructions online. I learned the basics of a python extension called tensorflow, which has built-in mathematical formulas to simplify all the tedious math that goes into the code. But I ran into a roadblock. The next step was basically to start writing my own algorithm, which I had no clue how to do. My mentor was very busy, and I couldn’t go to him every time something came up.

This is when I realized that the internet is only helpful when someone has already found a solution to my problem. Since what I was coding and researching was relatively new, I couldn’t find anything online to help me. My whole life I have felt that technology has virtually no boundaries, but in this instance, I could see the limits of the internet. I felt pretty powerful after my discovery, but at the same time, I was still stuck.

There was a time period of about a few weeks where I struggled to make progress. This lack of help from my favorite friend Google was killing my spirits and I felt like I was disappointing my mentor. Other than the lack of progress, the more irritating part of the situation was that I thought I knew more than I did. When I was going through the steps on the tensorflow website, I was basically copy-pasting and not thinking critically about the implementation of the code. I did not have the fundamentals and therefore fell apart when I had to be creative, making me realize that the internet is not the fix-all option for finding answers.

In the end, I ended up working through the issues with my mentor and finished my final project: running a dataset onto my own generative adversarial network. The experience was amazing, as I was able to see how business works in the STEM field. It is just  somewhat ironic that the place where I learn the limits of technology is one of the leading technological research companies in the world.