I bought an MX5 NC Automatic in May 2024 and have done approx 5k miles on sunny days exploring UK mountain passes and Bavaria.
For route planning I have been experimenting with ChatGPT (General AI program) for places that are new to me.
Method
For asking questions I have used the following -
“What are good roads and mountain passes in (Place name e.g. Lake District / North Wales / South Wales / Bavaria / Exmoor ) to enjoy in a small sports car?”
Results
As ChatGPT has the knowledge of these areas, from its analysis of most documents that have been written about them, it is able to give me a list of recommended roads with reason for choosing them.
The list is more comprehensive then any research I could do on my own. I then use the output with a map to plan my actual journey. The list has never been wrong!
that’s an interesting idea. I’m still one of the people who simply can’t find a practical (reliable) use for AI, when I’ve tried using it to produce stuff on subjects I know a bit about I find lots of factual errors, or I’ve ended up with useful text that I then end up editing so much I might as well have done it myself in the first place. But I will try this, thank you.
AI has some positive uses and also some significant negative effects (massive amounts of energy required to power it being one) and generally is not nearly as clever as media output would suggest. Not to mention ‘stealing’ of source data for the algorithms to mine and use to return an output.
Large companies that have gone all in are desperate for a return on investment where observations suggest that in the hype cycle (linked and a proven model for new technologies) we’re at the top of the ‘peak of inflated expectations’ and rapidly heading into the ‘trough of disillusionment’.
Algorithms feeding on stolen data sets can return interesting outputs, is it better than getting out an ordnance survey map and tracing roads and contours? That’s subjective, quicker yes, as gratifying?