I bought my phone for the non-phone features, sure I'll admit it. In fact, I never had a use for a cellphone until you could browse the net on one. When that came along, I was all in. For me it's not a parlour trick, it's the primary reason I have one.
Does anyone believe a modern Benz will be able to drive for 1,000,000 miles the way old ones regularly would? I don’t.
I think this has to do with the advances in material sciences and engineering. My theory is that parts in old cars were massively overengineered because the variation in quality was so high. If you wanted most of your cars to last 100,000 miles, you had to construct a car to last 1,000,000 miles. Nowadays, you can probably get away with constructing a car to last 200,000 miles.
Can someone with the right background comment on this train of thoughts?
The problem is more one of cost and efficiency. You can still run your care for 100k of miles (the 1M miles is fiction, imho). But when e.g. the engine is broken, you don't fix it anymore but replace it wholesale. Less hassle, guaranty on the new part by the manufacture and generally cheaper: The old engine gets sent to the manufacturer for refurbishment by people only doing this (and hence more efficient than your run-of-the-mill mechanic).
Furthermore you can't fix a car do-it-yourself-style anymore because these things have become way more complex. So running a car _way_ past it's economically sensible lifetime (if you'd factor in the labour cost) is simply no longer possible.
And even old cars totally overengineered do fail all the time (my brother has a bunch of them). But since it's usually just for leisure, labour cost doesn't count. And because the parts are simpler it can usually be fixed.
So I'd blame this on the massively increased labour cost/car cost ratio. Go to a country with really cheap labour and it all works like it did here half a century ago.
You bring up an interesting factor I didn't really take into account. It probably makes no economic sense to run a car for 1M miles considering the high costs of labour. The question is whether it is possible to run a modern car for so long with somewhat reasonable maintenance, or if it will die with a fatal engine failure.
We will see in about 15 years, whether the modern taxis of today still serve as taxi in lesser developed countries.
Fun anecdote: I recently sat in an old BMW taxi (I guess early 80s). The mileage indicator showed about 50,000 km, which is way to little for any taxi. The indicator had 6 digits. My guess is that it reached 999,999 km, overflowed and started again at 0.
"I make no secret of the fact that I think real technological progress has slowed in many fields, possibly even reversing itself." states the author at the start. That's very hard to defend.
However, let me boldly agree in that I think this is true for qualitative change, but not quantitative. I think quantitative changes are rapidly happening, and that's because it's very easy to generate better answers from a set of pre-existing knowledge. Computers (plus their engineers) are very good at that. It's the whole premise of the Semantic Web.
The hard thing to do is generate a better set of existing knowledge, and then curating it well. As (perhaps mis-used but I'll use it anyways) Picasso said "Computers are useless. They can only give you answers."
It'd be interesting to know the price drop over time of the cars. I'm sure that cars nowadays are more cheaper than back when they could get a million miles, so the comparison doesn't work.
> Revolutionary jets like the SR-71 or the 747 took months to design.
Yes, and without computer assisted modelling, when they wanted to try out something - say, reducing trim drag by moving the center of gravity further aft - they had to just try it out in the field. That experiment killed one pilot, and it's a miracle the other survived. What caused that death? Engine unstart. Why was unstart a problem on the SR-71? Because its analogue computers couldn't always keep up.
All in all a spectacularly bad example, CAD would have saved lives, and despite his ranting it DID have a computer on board, and when that was replaced with a digital computer system it increased the reliability and safety of the plane.
The author's rant is mostly crotchety nonsense, but it was worth it for the link [1] to Harry Plinkett's hilarious and spot-on critique of Star Wars Episode III.
11 comments
[ 3.9 ms ] story [ 36.5 ms ] threadI think this has to do with the advances in material sciences and engineering. My theory is that parts in old cars were massively overengineered because the variation in quality was so high. If you wanted most of your cars to last 100,000 miles, you had to construct a car to last 1,000,000 miles. Nowadays, you can probably get away with constructing a car to last 200,000 miles.
Can someone with the right background comment on this train of thoughts?
Furthermore you can't fix a car do-it-yourself-style anymore because these things have become way more complex. So running a car _way_ past it's economically sensible lifetime (if you'd factor in the labour cost) is simply no longer possible.
And even old cars totally overengineered do fail all the time (my brother has a bunch of them). But since it's usually just for leisure, labour cost doesn't count. And because the parts are simpler it can usually be fixed.
So I'd blame this on the massively increased labour cost/car cost ratio. Go to a country with really cheap labour and it all works like it did here half a century ago.
We will see in about 15 years, whether the modern taxis of today still serve as taxi in lesser developed countries.
Fun anecdote: I recently sat in an old BMW taxi (I guess early 80s). The mileage indicator showed about 50,000 km, which is way to little for any taxi. The indicator had 6 digits. My guess is that it reached 999,999 km, overflowed and started again at 0.
However, let me boldly agree in that I think this is true for qualitative change, but not quantitative. I think quantitative changes are rapidly happening, and that's because it's very easy to generate better answers from a set of pre-existing knowledge. Computers (plus their engineers) are very good at that. It's the whole premise of the Semantic Web.
The hard thing to do is generate a better set of existing knowledge, and then curating it well. As (perhaps mis-used but I'll use it anyways) Picasso said "Computers are useless. They can only give you answers."
Yes, and without computer assisted modelling, when they wanted to try out something - say, reducing trim drag by moving the center of gravity further aft - they had to just try it out in the field. That experiment killed one pilot, and it's a miracle the other survived. What caused that death? Engine unstart. Why was unstart a problem on the SR-71? Because its analogue computers couldn't always keep up.
All in all a spectacularly bad example, CAD would have saved lives, and despite his ranting it DID have a computer on board, and when that was replaced with a digital computer system it increased the reliability and safety of the plane.
Here's the story from the survivor if anyone's interested: http://www.alexisparkinn.com/sr-71_break-up.htm
> The next day, our flight profile was duplicated on the SR-71 flight simulator at Beale AFB, Calif. The outcome was identical.
This seems to belie your argument.
[1] http://www.redlettermedia.com/plinkett.html