I hate to rain on anyone’s parade, but I have to say that in my reading about the Pittsburgh self-driving car test, it sounds like the cars are almost comically not ready to go, and by that I mean, way not ready to be fully self-driving.
First, article from *Business Insider.* They all have two people in them. Examples of what the cars have trouble with: 1) knowing what to do when they have to stop, wait, and start again for unexpected obstacles (such as geese crossing road), 2) bridges (not enough cues to tell them what to do. Isn’t this a biggie?); 3) humans not obeying traffic rules (driver in right lane turning left); 4) possibility of snow obscuring car’s navigational cues and its sensors. This in Pittsburgh, where there is a lot of snow.
Uber’s self-driving cars are impressive — but there’s still a lot they can’t do
Last week, Uber rolled out its self-driving cars in a pilot program in Pittsburgh so that select users could experience…www.businessinsider.com
Second, article from *Slate* — http://www.slate.com/blogs/future_tense/2016/08/18/uber_driverless_cars_in_pittsburgh_actually_have_drivers.html — making the analogy that the cars right now are self-driving in the way a bicycle is. Yes, they can propel themselves, but only if someone is there to grab the handlebars if something happens.
I agree with everyone that self-driving cars would be great. Just the idea that we could replace so much land used for parking lots with parks, actual parks, and have a fleet that services people like taxi’s do not is great for our pocketbooks and great for the environment.
But to tell the truth, this just seems to me like the cart is far, far, far before the figurative horse, and I do mean whatever metaphoric pun that summons up.
The difficulties in this beta test might take years to debug. We have no idea how expensive it will be to retool the issues manifested in the Pittsburgh test.
When you add to those issues the fact that Americans love the *autonomy* of their cars and the privacy, it is by no means clear to me that the self-driving car will become a thing even if they are workable. There will be cultural issues surrounding their genuine acceptance.
Look how long it’s taken us to get hybrid cars in wide use, or alternative fuel cars. Forever. We’re not there yet.
We might get cars with self-driving capability or more computers than already exist. But I doubt we will have the fully robotic, “kick back and watch Netflix while it takes you places” vehicle people are talking about.
The whole thing reminds me of this: In the go-go 1990s, I worked on Wall Street. Many companies would seek to wow analysts with their cutting-edge technology. One I remember was voice-to-data technology. A company held a huge conference touting this, that they had a software and system that could convert anything said to a written document. Automatically.
Except when you actually looked at the prototype, it wasn’t really capable of doing it. Yes, it could record a spoken sentence like “I would like a class of water.” But what it spit out in exchange was something like “I need to watch my weight.” Not enough commonalities to be remotely the same thought. (And the start-up had received thousands of dollars for developing this, mind you.)
And, decades later, voice to text technology is still really not that usable.
For all the very real innovative strengths of tech, overstating and hyping the coming thing is also part of that world.
As I say, I’m not one for raining on parades. But the idea that we’ll have these in 5 to even 10 years? I’d be willing to bet money, Billfolders, money, that we won’t.