Block is a student majoring in political science at California Polytechnic State University in San Luis Obispo. He’s originally from Encinitas.
AI, or artificial intelligence, is really cool and really scary. Not scary like “the terminator,” but because AI is increasingly replicating things we typically define as “human.” A few years ago, near-instant text and image generation sounded impossible. Today, posts showcasing ChatGPT and Dall•E 2 are flooding social media. As somebody who grew up in a digital age, I have to imagine this is a similar feeling to the early days of Google and similar search engines. Yet I believe the repercussions of advanced AI and machine learning mean far more for our society. Let’s consider implications in three areas I am interested in: entertainment, education and elections.
“The use of AI in the entertainment industry has the potential to bring about both positive and negative changes. On the positive side, AI can be used to assist writers experiencing writer’s block, providing inspiration and ideas for new stories and characters. However, there is also the potential for AI to be used to generate large amounts of low-quality content with minimal effort. AI algorithms can be trained to produce content that is designed to be as attention-grabbing as possible, regardless of whether the content is of high quality or not. One concern is that AI can be used to create content based on the work of artists and filmmakers without their consent, prompting ethical and copyright issues.”
But would anyone even want to read AI-generated content?
Hate to break it to you, but the quote above was written by ChatGPT. All I asked it to do was to create the “entertainment” section of an article given the title and some key points. This is not to suggest that humans will lose their place in entertainment creation. Pieces with more emotion and contextual analysis will be safe for some time. But content like sports updates, traffic alerts and weather forecasts may become entirely run through a large language model like ChatGPT, especially in text-based media.
If search engines made cheating for students easier, ChatGPT essentially hands them pre-cooked answers on a platter. It spit out well-written, detailed answers to any question I could think to throw at it (including from some old quizzes I had access to). But accuracy is ChatGPT’s biggest limitation for students at the moment. This typically manifests itself through incomplete, outdated and sometimes simply false information. Graders will have to rely on this, as there is no easy test to see if text is AI-generated at scale. Yes, tools exist, but they can be fooled (especially when mixing human and machine-generated content) and aren’t efficient to use for hundreds of assignments.
In fact, I’m certain that teachers across the world are reviewing machine-generated responses to their questions without realizing it.
While this seems like a hindrance to education, that AI can also be an incredible tool for learning. An informed chatbot can act as a personal expert for any student to consult. Try it for yourself: ask ChatGPT question after question about a niche topic. While it isn’t perfect, it has more than enough potential to radically change how we learn.
Elections are complicated enough already, so predicting AI’s impact here is a little more difficult. Still, it’s easy to imagine how well-funded national campaigns might utilize AI. Voter outreach can be dramatically changed with a chatbot modeled after a candidate who can give instant, curated and dynamic responses to voter questions. Humans will likely still have a role in creating connections over the phone and in-person, but AI can outreach quicker and possibly more convincingly than a human. Voters may feel more drawn to candidates if it feels like they’ve talked with them personally, even if they know they were talking to a bot.
On the flip side, negative fake interaction will likely skyrocket. Social media already faces a bot problem, and this will likely become worse as bots better imitate humans. What happens when thousands of s can be directed to share the same sentiment (positive or negative) about a candidate, but they are indistinguishable from real people? Furthermore, malicious bots spreading deepfake videos will only continue to skyrocket in of both quality and quantity. Methods to reveal deepfakes may evolve with it, but by the time they are called out, the damage may have already been done.
The AI world is both exciting and terrifying. New tools will open countless doors to reach heights we never thought possible — but they may also shut doors we thought safe. My predictions above are merely predictions, but I can say this with 100 percent certainty: AI technology will rapidly advance in the coming years and society will be forced to adapt. I sincerely hope we open more doors than we close in that process.
