Using Made Up Data To Train AI

Did you know that AI can be biased in its decision making? Me either.

In my Ethics class, we had to write a technical research paper as a group. I really wanted to do something on the ethics of Self-Driving cars, but my group decided to write about Synthetic Data for AI Training.

In writing this paper, I didn’t just learn a lot about what Synthetic Data was, I also learned how to work with people who had very different views. Sometimes we didn’t agree on what topics we should cover, or how we should edit the paper together. We also had to manage our expectations, especially when none of us wanted to write the Executive Summary, but it was due the next day.

One of the most interesting things I learned about Synthetic Data was that the US copyright law covers computer programs under literary works, and protects the “expression of original ideas” not the code itself. AI systems are considered an expression of an idea, and that can be protected under copyright.

You can read the paper here: Synthetic Data in AI Training