> “They don’t build stuff like they used too anymore”.
In fairness part (but far from all) of the reason we don't build stuff like we used too, is because we have much better understanding of material limits so we don't need to over-design structual elements like we used to.
How that understanding of material limits feeds into design obsolescence is where I do get annoyed though.
With everything we know now we could (obviously) build things better (for whatever criteria you chose) than they did 50 or 100 years ago, if that was the goal but it rarely is.
My favourite piece of civil engineering in the world (because it's nearby and I've ridden over it many times) was built in 1980 though (it's the exact same age as me) and its still awe-inspiring to look at.
The problem is that if a company gives customers (both b2b and b2c) the choice between a cheaper product with minimal guarantees vs. a product that is indeed better and can last longer, not enough customers will choose the better (more expensive) product to justify development cost (there are exceptions obviously).
Partly, companies are even forced if the lowest bidder for pre-defined specification gets the contract.
I believe there are more than enough engineers and companies that would like to build better and more durable products but demand just isn't there.
The problem is you can't trust anything/anyone any more... Brands are generally trusted until you get burned by one. And then they relabel or buy out another brand to burn.
I often make choices that are anywhere from 50% more cost to 200% if I'm relatively certain I'm actually getting a better product. Unfortunately that trust is usually limited and even then, you often don't have a good trust case.
There are also times where I look at industrial options for comparison, as that's often usually skewed towards more solid/maintainable production. Though again often means more expensive.
I tend to find the best value (my own experience) about 2/3 up the pricing scale for whatever products I'm looking at.
If I could find appliances in general with a 15+ year warranty (that didn't have weird loopholes), I'd be more inclined to buy. Unfortunately, they often don't cover interface boards, and even then they redesign in often incompatible ways each year and stop making the previous version entirely.
As someone who has spent most of their working life so far trying to make it easier to store data on materials, I'm not sure things have progressed all that much.
I do work with people from Boeing (and Airbus) but they are not yet using all the features of stuff I wrote 30 years ago.
I think it may be that brilliant engineers still exist somewhere, but I suspect they're less likely to go into aerospace than in the 50s, 60s or 70s. Heck or even the 40s.
Immediately after the war aerospace was red hot - there were countless countries and companies putting out new jet designs - both commercial and military. The competition and pace was intense and there was quite some cachet and public interest in aviation.
Since roughly, I guess, the 90s aerospace is rather settled and a very mature market. There's only a few major players left, and very very few new products; many of those are variations on the previous. Not nearly as much place or appeal for the brightest stars.
Firstly, from 1903-1968 aerospace was in a period of exponentially improving performance (range, speed, altitude) comparable to Moore's Law. But from the mid-60s onwards the rapid improvements stopped, as developments hit hard limits imposed by material science and thermodynamics: it's hard to get a material much stronger than carbon fibre or titanium alloy, it's ridiculously hard to go faster in air (drag increases faster than the square of speed: if you throw more energy at the problem via your engines it comes out in the form of friction-induced heating).
Secondly, there have been huge incremental improvements since 1968. Consider the 737 series. The 737-100 had a maximum range of 1540 nautical miles on 17.86 metric tonnes of fuel; today's 737MAX-10 can do 3850nm on 25.94 tonnes of fuel: it's 72% more fuel-efficient. It's also gone from 118 to 204 seats, so about the same upgrade is passenger capacity. The cargo capacity has almost doubled, too.
This being HN, the point of comparison should be an Intel-cpu laptop circa 2009 with an equivalent laptop today. The headline clock speed is probably similar (stuck around 2.4GHz), as is the number of hardware cores. The modern laptop probably has more RAM and more storage, but only by a factor of 2x or 4x. But it has double or triple the battery life, 4x or 8x the pixels on screen, a screamingly fast SSD instead of a rotating-platter drive, and a GPU that the OS can offload a chunk of work onto other than simply accelerated graphics. It looks disappointingly similar, but if you could give your 2009-self your 2019 laptop they wouldn't be in a hurry to give it back again.
There have also been huge improvements in overall safety, the 737MAX MCAS problem aside; compare accident rates today with accident rates in the 1980s, let alone the 1950s if you want an eye-opener. (Of 1010 Boeing 707s delivered, there have been 173 hull losses in accidents.)
As a brilliant star with a newly minted engineering degree I might go into aerospace pre-1968 to play with blank sheet developments, X planes and XB-70, Concorde, Harrier, SR-71, 707 etc etc. Much the same with at engine makers - Merlin to Olympus and Pegasus. An absurdly long list, often with remarkably small teams. I might not want to take my 2019 vintage degree into a world of incremental improvements at one of the few global giants. Even if those improvements are of enormous significance over the years. It doesn't feel earth shaking any more, Farnborough isn't showing astounding free PR every year, just steady incremental progress (mostly). No more flying bedsteads.
Same goes with the laptop analogy actually - the fun was probably in the earlier years than now with new model every year with an extra few percent on the numbers, or slimmer. Green fields are usually more fun, and when they succeed, more satisfying; even with the odd Osbourne effect or total failure.
Were I that newly minted genius, I'm not sure where I would park myself today. Space X for some, perhaps. Probably not Boeing or BAe. I'm sure there's plenty of good (enough) engineers at all of them still. Working as one of thousands. That's not so alluring.
> 1010 Boeing 707s delivered, there have been 173 hull losses in accidents
I forgot the point I was trying to make while giving the background!
It's not that the engineers of yesteryear were brilliant while today's engineers are second-rate: it's that what they were doing back then was in some respects easier. (And we remember the good planes: as often as not they invented utter turkeys: look at the history of the Supermarine Scimitar for example — https://en.wikipedia.org/wiki/Supermarine_Scimitar — TLDR: yes, it's military, but to lose 51% of your aircraft in 9 years of peacetime service takes some doing. And then compare with the F-35, two crashes out of 308 airframes in service over a 5 year period, even though it's a vastly more complicated and higher performance aircraft.)
And more equally distributed ones: there are not many easy solutions anymore that are pareto improvements, ie do not have downsides elsewhere.
For example, the door to the cockpit is locked now to keep terrorists out - which allows a suicidal co-pilot to fly the machine into a mountain undisturbed. (Having said that, a suicidal pilot can take down the machine one way or the other, so maybe locking the doors was a pareto improvement.)
there are very few brilliant engineer of yesterday whose job role today is to train the next generation of brilliant engineers so that their vast experience doesn't get lost in time; the brilliant engineers of today have to relearn their experience walking the same mistakes, so that even companies that pays those premiums for senior engineers will have to face the same generational problem as everyone else, because they don't want to waste highly expensive professional to something as mundane as training on the job.
>We haven’t gotten stupider as a species in 30 years
I'm honestly not sure that's true. On top of anecdotes (in my grandfather's generation, the norm was to fluently speak 3-4 languages), go ask any STEM prof. with 20+ mileage about the general math, science and logic literacy of students nowadays.
> On top of anecdotes (in my grandfather's generation, the norm was to fluently speak 3-4 languages)
I'm not aware of anyone in my grandfather's generation that can speak more than one language fluently. The fact that my grandparents grew up on farms on the edge of the Dust Bowl may have something to do with that, but I rather suspect you're attempting to compare a tiny subset of the population in the past against the majority of the public today.
There's also the prevailing management theory that "all staff can do the same job" -- I've seen this in various perverted and butchered variants of Lean and Agile.
The theory is sound -- at least two engineers on a project should know how a given functional unit works. It's a practice for lessening the impact of "bus factor", where you have one engineer everyone relies on because they're "the one person in the whole company who understands it".
The issue is when this kind of thinking reaches management, and you see someone being pushed into a role when their skill set is radically different to what's needed.
And then there's surprise and finger-pointing when things (inevitably) go wrong...
Indeed, and I'd add that we are going cheap because we can, thanks to computing power.
As a matter of fact : computing tools make a lot of things easier and cheaper (which is good). But the setback is a tendency to "laziness" in the whole process...
As people say “They don’t build stuff like they used too anymore”.
First we started going cheap on the materials. Now we are going cheap on the engineers too.
Why pay a senior “brilliant” engineer $$$$ when a fresh out of college “average” engineer can write the “same piece of code” for $$