NavList:

David C, you wrote:
"What about Uber drivers whose livelihood depends on GPS?"

Thank you for replying. Yes, I would wonder about their economic cost, too, and part of the reason I posted that story about the economic costs of planting rows of seeds is because I'm skeptical of it. There are other economic activities affected by lost GNSS signals, like the drivers you mention. And I'm not convinced that seed cannot be planted without satellite signals! The source of the story is a popular astronomy magazine, and it could just be anecdote.

You also wrote:
"I recently took an Uber on a rather long route. How frustrating. Several times we passed an intersection where, if I was drioving, I would have turned off to go directly to my destination. But the uber driver followed a long round about route with heavy traffic. He obviously did not have a clue where he was going and relied on his phone on the windscreen."

There are lots of "talking points" around this story :). Of course the GPS/GNSS signals merely provide positioning information: lat, lon, etc. That we all know, and I'm only getting out of the way. Ah but the mapping and driving app...! That's the real intelligence here, and of course we colloquially call this "GPS", too. As in "The GPS told me to turn left, but the road was closed..." or "The GPS told me to take Highway A but County Road 5 would have been faster..." That's not really GPS but we've all gotten used to this terminology.

Are the driving apps reliable? Not perfectly. But I'll say this: I always trust my driving app first. I use Google Maps, almost exclusively. In the case you describe, it's possible the driver was using a poor app (out of habit), or possibly one that was not optimized for New Zealand driving (there might even be privacy limits on data in NZ), or maybe the driver had activited some settings that always prefer major highways. Those of us who have been driving for decades have this feeling that there are shortcuts, and there are tricks to avoid traffic jams, and our decades of experience make us experts in these things. But experience with a good driving app and testing how it figures things out have convinced me that the app is almost always doing it right, so long as there are no weird settings activated. Yes, crawling along in heavy traffic on a major highway is painful, but it's almost always faster than the shortcuts that we can imagine. And the app does occasionally suggest shortcuts that are clever to the point of being devious --tricks like get off a highway at this exit, ride down the parallel access road for a mile, then hop back on. Now that's a guilty pleasure!

I've been watching these driving apps develop and become more efficient for thirty years (yikes, but yes, since about 1994). In the early days, you could sit and watch them sniff out different cross-country routes. After a minute or more of processing, the app would display its best guess at a fast route. In those days the analyses were driven by little more than road mileage and speed limits. With live, nearly instantaneous traffic conditions, the apps can now optimize in real-time, and the results are hard to beat. This is tech at its best, as I see it.

Driving apps and their route-finding algorithms are so finely honed and so widely used that I expect we're finally entering an era where feedback will develop between the apps and the traffic. If there's traffic on route A, re-direct everyone to route B... Oh, hey, now there's heavy traffic on route B, so send everyone to route A... No, no, it's better on route B! ...and so on. I'm sure that there are people in the industry and in academia studying this very problem, so perhaps it will be resolved as quickly as it appears. It is in fact, quite close to the original problem of "cybernetics" as originally defined (1948? by Norbert Wiener). Like a ship with an autopilot over-controlling the rudder to steer a steady course (kybernetes supposedly Greek for "helmsman"), leading to oscillations as the course steered overshoots to the intended course, app-controlled traffic could develop artificial oscillations induced by over-control.

You concluded:
"I have just thought......   do the black cab drivers in London stll take the "knowledge" or do they use gnss?"

I know that they still publicize "the knowledge" as their reason for existence. Do they eschew apps? I don't know, but I certainly hope not. A knowledgeable, intelligent driver with good communication skills aided by an app or two should easily beat an un-communicative driver applying only old-school "knowledge" and will usually beat a driver using only an app. I have a recent single-case experience that is at least a good travel story, but from Cambridge not London. After an unremarkable train from King's Cross to the Cambridge train station, I knew how to get to the little hotel where I was saying, and it was only about 15 minutes on foot (I walked it when departing, no problem), but I was beat, and as I stepped out of the train station, there was a line of cabs. So I gave in...

My cab driver in Cambridge spoke very little English. I could not guess his native tongue, but something southeast Asian, maybe Burmese at a guess. He did not know my hotel --bad sign, but it was a small "guest house", so I could forgive that. I did not have the address of my hotel right in front of me, but I knew it was close to the "Scott Polar Museum", which is well-known, so I asked him to take me there. He had never heard of it (!!) and he pulled out a little guidebook and started digging through it. And here's where we find a flaw in the whole "cab" or even "Uber" system in the modern world. By now, I had Google Maps out, and I could see the route we needed to take. I suggested certain road names. He clearly couldn't understand me, so we were reduced to "that way! that way!" and "left here please!" from me in the back seat. This all worked in the end, but it was more trouble than it was worth. And neither a Cambridge equivalent of "the Knowledge" (he had none beside his little book) nor my modern app (stuck as it was in my hands) resolved the problem.

Frank Reed