Or maybe not AI, technically, but something AI-like. And maybe it’s been done already, it’s been some years since my last road trip.
Scenario: I was on a long road trip – multiple days. Driving North back to Minnesota, it was late at night, and I needed coffee. I was following either Google Maps or Apple Maps on my iPhone, or more likely it was the car’s nav system, not because I needed help navigating, but because I found it convenient to be able to glance over and see a reasonable estimate of how much time was left to drive. I was fighting fatigue. It was time to stretch, get fresh air, and most importantly, grab some coffee.
So I asked whatever navigation/map about coffee.
It gave me options in all directions, half of which were behind me, most of which were a long diversion from the route I was already on. Horrible options.
What was needed was something close to I-35, North of my current position, to the East (i.e. right side) of the interstate.
See, the problem was that whatever navigation assistance I was using was not also taking my route into consideration when doing the search. It wasn’t trying to reduce the diversion. It wasn’t trying to reduce left turns from a driveway onto a busy road. It wasn’t trying to save miles, time, or risk.
In a situation like this, a McDonalds 12 miles ahead on the right, close to I-35 is far superior to Starbucks, DunnBrothers, Caribou, or any other “good coffee” place that was behind me, a half mile into some town, within a mile of my current position. That’s the sort of context clues that would actually help people.
Most of the “AI” and LLM things that dominate the zeitgeist lately are not designed to help people, contrary to the hype. They are designed to make filthy rich people richer, at any cost to society. So the lies about what they can actually do, told to the ignorant, are of malicious intent.
