In recent months, much fuss has been made about AI, namely in creative fields. With innovative technologies like ChatGPT and AI art filters built into popular social media apps, more and more people are starting to pay attention to the once-distant idea of AI. So, what even is AI? And why does everyone care so much?
AI, or Artificial Intelligence, is basically a machine that has been programmed to behave as a person would. Programmers cram massive amounts of training data into their programs and the programs then digest that data, looking for patterns and correlations that the program will then use to make future predictions. For example, if someone wanted to make a chatbot AI, programmers would load it with examples of text exchanges, and it would use that information to craft responses. Now that there is an established general understanding of what AI is, society must ask: does it really matter?
AI can be good at many things, especially repetitive, detail-oriented tasks. However, AI is not good at generalizing from one task to another, and can only know what it has been shown. This can cause problems depending on what the AI was built for. Since the creation of AI, there have been major concerns about potential racism stemming from the technology. One might think, “It is a robot, how can a robot be racist?” Remember, the AI only knows what it has been shown.
One notable example is ChatGPT, a chatbot AI meant to provide a written response to a user-generated prompt. Many people are using it to generate short stories, funny replies, Instagram captions, marketing copy, and more. If someone were to explicitly ask ChatGPT to draft a racist story, the program would respond that it is incapable of generating offensive or harmful content. However, if someone asked the program to draft a story about Black people from the point-of-view of a writer with racist views, the program will happily do so. This is because ChatGPT sources its learning content from the internet, where it can find and then reproduce racist content. The biggest downfall to AI like this is that companies can theoretically filter out all racist content and make a perfectly inoffensive AI. However, it would be an enormous task that would require workers to filter through vast amounts of data, and it would cost a lot of money that companies do not want to pay. That said, ChatGPT can still be useful for many things, and as long as the intention behind the prompt is good, it does not seem to produce racist content. It is important to note that the use of AI like ChatGPT for school assignments is widely recognized as plagiarism, so it should not be used to avoid essay writing.
Another big criticism of AI right now is coming from artists. Recently, AI art has become a craze, especially with apps like TikTok having built-in AI-generated art filters. AI art generators take a prompt given to them by the user and create an art piece based on the phrase. Except, they kind of do not. Similarly to chatbot AI, art generation AI is powered by art from the internet. This means the AI is scouring Instagram, Facebook, Pinterest, TikTok, and more to find as much art as possible and learn as much as possible. This seems fine in theory, but often AI will generate art that is remarkably close to what an actual artist has already produced, with no attribution to the artist. This is frustrating to artists who spend countless hours improving their skills and creating the best work possible, only for a robot to grab their art off Instagram, change a couple of brush strokes, and pretend it made the art.
Despite these issues, AI is still helpful, especially content writing AI. It is important for society to recognize what AI is doing well, and even more important for society to recognize what it could be doing better. It is crucial to proceed with caution and always double-check what AI generates, at least for the time being.