Since the 1990’s the internet has provided a portal to the world of information. We could simply type a question or topic into a search engine and be provided with a long list of possible resources on that topic. It was then up to the researcher to analyze and evaluate those resources to determine which were useful and which were not. We often refer to this skill set as critical thinking. Merriam Webster defines Critical thinking “as the act of applying reason and questioning assumptions to solve problems, evaluate information, discern biases, etc.” This was a quantum leap beyond the former process of using physical references such as encyclopedias, journals and books to research a topic. Now the entire body of knowledge on a subject or question could be obtained in one search.
Artificial Intelligence Makes Mistakes
With Artificial Intelligence (AI) embedded into all our online software applications, current internet search results now lead with an AI summary which offers references for those who are willing to make the effort. It turns out that trusting AI may lead to mistakes, bias and even what are referred to as “hallucinations”. AI is built upon Large Language Models (LLM’s). These AI systems analyze massive amounts of text data to understand and generate human language (Stanford University). However, depending upon the topic, AI may provide incorrect or even fictitious information 15%-50% of the time while only 27% of users reliably fact-check the summaries. For some areas of research such as legal inquiries, it was found to contain errors up to 88% of the time (SQ Magazine).
Critical Thinking requires Practice
A concern regarding the cognitive development of youth is that the use of AI often eliminates the need to develop critical thinking skills that were once honed through analysis and evaluation. While it is a time-saver to obtain a summary in 30 seconds, the time taken to read the references and cross-check information would have provided a valuable opportunity to learn new concepts and build those critical thinking skills. Additionally, placing trust in AI summaries without factchecking may lead to misunderstanding and the sharing of incorrect information. Even universities are concerned about AI reliance leading college students to “cognitive surrender” where they turn to AI “at the first hint of struggle” (MIT study quoted in New York Times).
Adults must be engaged
Our role as educators, mentors and parents is to be engaged with our children’s AI use and to provide guidance on how to appropriately use AI generated results. Exercises such as fact-checking a simple query can be enlightening. Teaching youth how to form better prompt questions can lead to better quality results. Age-appropriate questions about how AI should or should not be used are useful ways to get the conversation started. The lines between generating work and editing work can be blurry and children need tools to help them make those ethical decisions. Talking through simple scenarios in group settings can help shed light on the varying ways that peers may view the use of AI.
For those interested in tackling this topic in a 4-H club setting, we have 4-H scenario cards that can be used to direct a conversation about acceptable AI use. 4-H club leaders are invited to borrow the cards or request that we lead the activity at an upcoming club meeting. If you would like to learn more about the Orange County 4-H program, please visit our web site, contact us at orange4h@ifas.ufl, or call 407-254-9222.
1
