You must log in or # to comment.
The trouble with the railways comparison is that after investing tons of cash the railways were built. With AI the GPUs have no value after 6 years (if that). So the investment must continue forever. It’s madness.
The other trouble with the railways comparison is that trains actually work and can generate a profit for their owners.
With AI the GPUs have no value after 6 years
What? GPUs don’t age. They might get old technologically wise, but they don’t just… die. The silicone chip itself doesn’t care about age.
It’s not that they don’t technically work. It’s just they’re no longer efficient compared to newer versions that can do more with less power. So to remain competitive you need to upgrade otherwise your cost to execute a model is too high.
Hyperscalers used to write GPU’s down to zero value after three years, over the last couple of years they’ve all increased this to six.
Nullshit i wrote.
It’s not actually the transistors that break down in flash memory. Flash memory works by storing charges in what is effectively a grid of capacitors, and in order for the data to remain stored, the insulating oxide layers in the cells need to be preserved. Every time a cell gets written, a charge is forced through the insulation with high voltage, and this degrades the insulation. A single flash cell might only have a few 1000 writes before this insulation goes bad and it no longer holds data. Modern SSDs have wear levelling techniques to make the drive as a whole last longer.
Transistors on the other hand don’t have any inherent degradation that I’m aware of other than external factors like corrosion. The first thing that’s likely to die on a GPU is the electrolytic capacitors in the power filtering electronics, which have fluid in them that dries out over many years.
I doubt the transistor on a GPU wafer break after 100k cycles, as they run at gigahertz frequencies, some cycle billions of times a second.
Transistors break after 100000 cycles.
SURE, BUDDY.
I’m not an expert but I was under the assumption that electronic components (including silicon chips and their internals) will age and give out on the decade timescale
Check out “References” part here
They picked the wrong history, in my not so humble opinion. The AI situation looks more like the dot-com bubble, recycled.
Railways? Good example of tech abandoned in favor of something else.
Laughs in European
Huh? Rail handles about 40% of long distance freight in the US.
But I want to know the tech that will replace 60 % of AI.
Maybe quantum dunno
Like, the automobile? It looks like the boom in the UK they were talking about was in the 1840s.
https://en.wikipedia.org/wiki/Railway_Mania
Railway Mania was a stock market bubble in the railway industry of the United Kingdom of Great Britain and Ireland in the 1840s.
There were primitive automobiles earlier, but the mass market automobile didn’t come around for a long time after that, and then it’ll have taken longer to get substantial marlet penetration.
searches
https://www.bbc.com/news/uk-42182497
It runs a bit off the edge — I don’t know how far back they had licensing and mileage data.
But extrapolating from those lines, I’d guess that annual distance traveled in the UK in autos on roads surpassed rail only in the 1940s or so, about a hundred years later.
That’s probably outside the investment horizon of people investing in the 1840s — in evaluating whether an investment is worthwhile, they won’t be considering returns a century hence.
That being said, it is possible to maybe consider freight rail, and it’s possible that that works out differently. The US doesn’t use much passenger rail in 2025, but it does do quite a bit of freight rail; the two can be decoupled.
EDIT: It can’t be too much earlier that road traffic could have risen, though, since mass-market motor vehicles weren’t much earlier than that.
I mean a hundred years is not much for a technology your government decides to build society around.
I mean, in that kind of timeframe, there were pretty major shifts in transportation.
For a long, long time, ships up rivers and along coasts was the way serious transportation happened.
Then we had the canal-building era in the US. I assume that the UK did the same.
searches
https://en.wikipedia.org/wiki/Canal_age
Technology archaeologists and industrial historians date the American Canal Age from 1790 to 1855[1] based on momentum and new construction activity, since many of the older canals, although limited by locks that restricted boat sizes below the most economic capacities[b], nonetheless continued in service well into the twentieth century.[c]
By 1855, canals were no longer the civil engineering work of first resort, for it was nearly always better—cheaper to build a railroad above ground than it was to dig a watertight ditch 6–8 feet (2–3 m) deep and provide it with water and make annual repairs for ice and freshet damages—even though the cost per ton mile on a canal was often cheaper in an operational sense, canals couldn’t be built along hills and dales, nor backed into odd corners, as could a railroad siding.
So that was maybe sixty, seventy years before rail was really displacing it.
EDIT: I guess what I’m trying to get at is that I don’t think that rail had a uniquely short era where it was the prime, go-to option compared to other transportation technologies…and I don’t think I’d say that the golden era was short enough to make the technology not a worthwhile investment, even if it was later, in significant part, superseded. A hundred years is a long time to wait around without engine-driven transportation, which would have been the alternative.
What’s the original link? The archive won’t load for me.
Edit: I got hit with a captcha that wouldn’t load in DDG mobile. I opened it on a desktop and I have it now.
How? Archive even loads in Dillo, in a no JS mode.
500 error, that’s service-side
Usually from too many requests for the server to handle
DDOSed again?
Probably not tbh, sometimes it’s just a hug-of-death
The main difference between the two is intention.
