Using ChatGPT
I thought it was time to give ChatGPT a serious go after all the hype - and some of the comments are hype which we always expect after some new technology becomes available.
I thought it was time to give ChatGPT a serious go after all the hype - and some of the comments are hype which we always expect after some new technology becomes available. But before we got there, here's a couple of things I found in the past few days.
First, the Atlantic Council offers us the results of their Global Foresight Survey which covers what the world will look like in 2023, their global risks assessment, and some 'snow leopards', underappreciated phenomena that we know about today. This is a survey of 'experts' and that influences the results because experts tend to answer based on their unchallenged assumptions and confidence bias about their expertise. And they tend to come from a constrained group of people. Read more about experts here. I am not going to go into any detail about these results except to say that I'm not sure there is a lot new in here.
Second is new reports from Shaping Tomorrow, using Athena, their AI robot, and ChatGPT. Â You will have to sign up for a free account from Shaping Tomorrow if you don't already have one, and then you will be able to access the various formats of the reports they provide. The latest one I got today is about Future Jobs - the link will take you to free slides summarising what the reports cover. I worked with Shaping Tomorrow for some years when I was a practitioner, and their work always has something of use - there is more functionality on the site too that you can explore.
Both of these groups focus on the doing of foresight, and Shaping Tomorrow is quite automated in the tools it uses. That's not necessarily negative, ofter it's very positive, but I still think that interpretation is still a human job. I'm happy to leave scanning and analysis to AI and even report preparation about the present, but I am still convinced that interpretation is a human job. Please disagree with me if you think otherwise.
ChatGPT
I started my chat with a pretty broad question and gradually moved to some more specific queries. I can see why people are going crazy about it's writing research reports, or for use in education. Or for writing in general.
The chat got quite long so I've provided it here as a PDF to download.
My reactions. The first one I had was similar to the one I had a few weeks ago when I tested ChatGPT - that it's very good for finding knowledge we know already. The more I worked on my questions, making them more specific, the more useful the answers seemed. The answers to Question 9 about assumptions were useful I thought, but I wasn't too impressed with its answers to my questions about scenarios (Question 11-13) - nothing wrong with them, but nothing new either, and I guess they could be a useful starting point in a scenarios exercise as the present scenario.
I would be, of course, interested in what you think about this chat and whether or not the answers are useful, interesting or another word that you can choose. Post your thoughts on the post comments on Foresight in the Present or on LinkedIn or Facebook.