The social media account "BrainyBotsClub" is a platform dedicated to educating high school students about AI, specifically how it works, what it does, and how they can use it in ways that allow them to still think critically. Our goal for our social media page is to help high school students navigate the growing world of AI. This will be achieved through education of the various types of AI, understanding of AI hallucinations, and increased AI literacy.

MAIN GOALS!

1. Help high schoolers understand that artificial general intelligence (AGI) does not exist, rather we have narrow AI models

2. Understand AI hallucinations

3. Educate about learning AI literacy and how that can help with future usage of AI

AI hallucinations are instances where an LLM creates information that is not factual and does not correspond to reality. AI has limits to its knowledge and training inputs, leading to these hallucinations.  

This is somewhat similar to human forgetting. Forgetting is an essential and adaptive function of human memory, allowing us to focus on relevant information and update our knowledge. This process is primarily caused by ineffective coding or cueing. These failures are observed within the different memory systems, including the working and long-term memory systems. 

Working memory (WM) is like a mental workbench; it's a small, temporary system that holds and actively processes the information we need right now to perform complex cognitive activities. This includes understanding language, acquiring new knowledge, and solving problems. Information is quickly forgotten from WM due to its limited capacity and rapid decay.

Long-term memory (LTM) is the system responsible for storing vast amounts of information that can be retained for extended durations, potentially spanning a lifetime. LTM is typically forgotten due to a retrieval failure where the memory is present but temporarily inaccessible. 

When we look at these ideas together, we can see that AI hallucinations are essentially the computer version of human forgetting. In people, forgetting often happens because the right cue isn't there to bring a memory back, or because other memories get in the way. In a similar way, an AI hallucinates when its system pulls up the most statistically likely answer rather than the most accurate one. Both the human brain and AI models are built to work efficiently and to generalize, meaning they focus on capturing the overall idea rather than storing every detail perfectly. As a result, both human forgetting and AI hallucinations show the same basic trade-off: being fast, flexible, and adaptable sometimes leads to occasional factual errors.