Media Explained — Misinformation: Generative AI
Dr Aimee Hourigan:
Remember that story about cat discos? Well, what would it mean to you if I said that it had been artificially generated? Artificial intelligence presents new and complex challenges to the way we interact with media and information. Let’s break it down.
Artificial intelligence is when computer systems or machines use human-like processes to solve problems, make predictions and complete tasks. Generative artificial intelligence takes this process of machine learning one step further by accessing huge amounts of information and using a special machine learning process known as neural networks. The AI system learns how to create its own digital content like new images, audio, video, or text. You might have heard the term ‘generative AI’ used to describe how certain types of misinformation are made. Deepfakes, for example, are images, audio recordings, or videos of real people that have been manipulated by these AI processes to make the person say or do something they haven’t actually said or done.
It used to be the case that we needed access to some pretty high-tech digital tools to create this type of content. Now though, these tools have become a lot more accessible and people can easily and quickly create and share artificially-generated content. This AI slop is quickly taking over the internet on social media. It’s often created and shared as engagement bait, fake and manipulated content designed to grab our attention and our reactions so that the account sharing it can earn money and build its follower base. The account posting the content might not even be real itself. It could be a bot, an automated software program designed to mimic the behaviour of a human user.
Bots are often used to spread misleading content about hot button issues like politics or health. When people want to try and convince others to support their agenda or view, using AI systems to share information doesn’t always come from a bad place. Government agencies, health organisations and businesses often use AI tools like chatbots and virtual assistants to help quickly give people answers to specific questions. A government agency, for example, might use a chatbot as a virtual helpline to direct people towards different online services. A business might use a similar system to provide details about their products or sales.
These types of AI systems aren’t perfect though, and interacting with them might lead to negative or harmful outcomes. If the information they provide is an AI hallucination information that looks accurate, but it’s actually false and made up, imagine if an AI chatbot told someone to use that false story about cat discos as legitimate evidence for a new law on pet ownership. When these systems make mistakes, we lose trust in them, the information they provide and the organisations that created them.
With all these challenges, it’s no wonder people are becoming more overwhelmed by the content they see online. But don’t worry. There are some techniques we can learn to escape misinformation. Just watch the next video to find out more.
