reading in hindsight: Jokester by Isaac Asimov
I like dipping back into the (personally) distant past and seeing what stories from those days have to say to us today. I picked up Jokester by Isaac Asimov earlier today; the core idea of it had stuck with me since I read it the first time, because I enjoyed the twist, but it was a smaller aspect of the story that caught my attention.
To add a bit of context, the news over the past few weeks and months have been filled with stories about ChatGPT3 and its cousins, and the impact that they might have on education (where I spend a chunk of my days), among other fields.
The aspect of Jokester that caught my attention was the core fact of the existence of the Grand Masters, such as Grand Master Meyerhof who drives the action in this story. This core fact is that there is a subtlety and a skill required to appropriately frame a question for a computer such as Multivac.
Or for ChatGPT3. I’ve not done a lot of playing around with it directly, but I know folk who have spent more time than perhaps is entirely wise. One thing that’s come out of my conversations with them, and the reading I’ve been able to do, is that, the output can depend a lot on how the question for such Large Language Models is framed.
And this brought me back to Grand Master Meyerhof; not actually the questions he asked, as interesting as they are, but that the skill of the dozen or so Grand Masters was to be able to frame the question.
Thinking of Large Language Models and their impact on education, just to focus attention, it seems that this provides a bit of direction for us to work to work with. These tools are not going away; the main question for us is, how to use them and more even than that, how to use them well.
I’m reminded here of the toolmaker koan, a novel by John McLoughlin, that I haven’t read for a long long time. But it does contain the basic question, is it always the case that a civilization will develop tools before it develops the ability to use them, and this strikes me as where we are at the moment.
We have developed some awesome and mighty tools. Perhaps these tools won’t allow us to answer Grand Master Meyerhof’s question as Multivac was able to, but they do pose this question for us. Will we have or develop the wisdom to use them well? Or won’t we? We will soon find out.
multijimbo 