.png)
This month we are discussing Chatbots from Word of Mouth by Michael Rosen. A really interesting chat about being aware of how Large Language Models (LLM) are created and the ability they have to persuade you that they are talking the truth. I know I have already stopped checking everything it gives me so our students will be even more convinced. I would love to hear what you think about this and what can school librarians do to help?
This is a fascinating podcast and would recommend everyone has a listen. I do believe that ChatGPT and AI are here to stay but we need to be aware of the problems in order for us to move forward. We have begun to move very quickly into the 'what can AI do for us?' mode along with 'how can it save us time?' but there is a darker side to this and is certainly worth focusing on too.
They talk about us being persuaded to believe all it says by the terminology used around it. Words like ChatGPT 'understands' 'listens' 'engages' and 'can reply', gives it a human feel which it really does not have. The other part I found useful was they talked about AI giving out 'non-information' where you could not check the truthfulness, accuracy or context of the words given out. You can be persuaded to think it knows by it getting many things right... you stop checking as you begin to believe what it is giving you is correct.
There is so much to take from this. I do think ChatGPT and AI has its uses and is here to stay but these warnings rang a bell for me that I could see myself being pulled in. How do we support our students on this journey?