Large Language Models and Nursing Homes

You know how Large Language Models require people to intereact with various platforms to build their knowledge base? Well I think it would be great if AI robots went into nursing homes at greatly reduced prices not only to help with physical labour, but also to ask patients about their memories of their lives. What the world looked like? What they had to do day to day to survive? Where did this happen? If we went back now to that place, would there be any evidence of those events? Who they loved and who they didn't. Encourage relatives to bring old pictures the AI robot can talk about with the patients.

Imagine gathering millions of these conversations together and processing them. Would it give us a new way to understand and study history? I think so. 

At the same time, these people will not be slotted away and left on some forgotten shelf of life. They will, in their own way, still be contributing to life on this planet. 

I think the more we study history, the more we will understand why things turned out the way they did. At present we allow a few clever historians to comb through documents and maybe talk to some key people who may remember those times. But those historians who have biases talk to key people who also have biases. And that's our version of history. Let's add another level of documenting our history. What will we find that we are missing out on presently?

At the same time, merely observing and intetacying with these elderly people might lead to medical and physcholical break throughs. Things we weren't looking for, but the AI sees a pattern developing and draws our attention to it. 

Talking to these elderly people who are at the end of life may also teach the AI a different perspective on life. At various stages of life we see the world differently. If one age group is over represented in these large language models, what does that say about the AI's prespective on life.