I suspect you'll see a few people "win" or strike it rich with AI, the vast majority will simply be left with a big bill.
It turns out that boom-and-bust capitalism isn’t great for building something that needs to evolve over centuries.
Perhaps American AI efforts will one day be viewed similarly. “Yeah, they had an early rush, lots of innovation, high valuations, and robber barons competing. Today it’s just stale old infra despite the high-energy start.”
For what it's worth, I like traveling by train and do so whenever I can, but I'm an outlier. Most Americans look at the travel times and laugh at the premise of choosing a train over a plane. And when I say they look at the travel times, I don't mean they actually bother to look up train routes. They just know that airplanes are several times faster. Delays suffered by trains never get factored into the decision because trains aren't taken seriously in the first place.
Humans are also freight, of course. It is not like the rail companies really care about what kind of fright is on the trains, so long as it is what the customer considers most important (read: most profitable). Humans are deprioritized exactly because they aren't considered important by the customer, which is to say that the customer, who is also the freight in this case, doesn't really want to be on a train in the first place. The customer would absolutely ensure priority (read: pay more, making it known that they are priority) if they wanted to be there.
I understand the train geeks on the internet find it hard to believe that not everyone loves trains like they do, but the harsh reality is that the average American Joe prefers other means of transportation. Should that change in the future, the rail network will quickly accommodate. It has before!
You are comparing USA today to the robber baron phase, whose to say China isn't in the same phase? Lots of money being thrown at new railroads and you have Chinese leaders and best and management leaders chasing that money. When happens when it goes low budget/maintenance mode?
Nonsense. The US has the largest freight rail system in the world, and is considered to have the most efficient rail system in the world to go along with it.
There isn't much in the way of passenger service, granted, but that's because people in the US aren't, well, poor. They can afford better transportation options.
> It turns out that boom-and-bust capitalism isn’t great for building something that needs to evolve over centuries.
It initially built out the passenger rail just fine, but then evolution saw better options come along. Passenger rail disappeared because it no longer served a purpose. It is not like, say, Japan where the median household income is approaching half that of Mississippi and they hold on to rail because that's what is affordable.
This is so misguided view... Trains (when done right) aren't "for the poor", they are great transportation option, that beats both airplanes and cars. In Poland, which isn't even close to the best, you can travel between big cities with speeds above 200km/h, and you can use regional rail for your daily commute, both those options being very comfortable and convenient, much more convenient than traveling by car.
What gives you the idea that rail would be preferable to flying for the NYC to LAS route if only it existed? Even as the crow flies it is approximately 4,000 km, meaning that at 200 km/h you are still looking at around 20 hours of travel in an ideal case. Instead of just 5 hours by plane. If you're poor an additional 15 hours wasted might not mean much, but when time is valuable?
Why would you constrain the route to within a specific state? In fact, right now a high-speed rail line is being planned between Las Vegas and LA.
But outside of Nevada, there are many equivalent distance routes in the US between major population centers, including:
Chicago/Detroit
Dallas/Houston
LA/SF
Atlanta/Charlotte
Passenger train service between Chicago and Detroit already runs 110 MPH (180 KPH) for 90% of the route, using essentially the same trainset that Brightline plans to use. They’ve got a rail tunnel underneath the Detroit River and had been planning to make the route be Chicago/Detroit/Toronto with Customs/Immigration happening on-board the train as it crosses the border. Ontario needs a lot of rail upgrades to make the Canadian leg run at those same speeds or higher.
They’ve made a lot of investments since the 1990s. It’s much improved, though perhaps not as nice as during the golden years when it was a big part of the New York Central system (from the 1890s to the 1960s they had daily trains that went Boston/NYC/Buffalo/Detroit/Chicago through Canada from Niagara Falls to Windsor).
During the first Trump administration, Amtrak announced a route that would go Chicago/Detroit/Toronto/Montreal/Quebec City using that same rail tunnel underneath the Detroit River. It was supposed to start by 2030. We’ll see if it happens.
Right now and since 1979!
I'll grant you that people love to plan, but it turns out that they don't love putting on their boots and picking up a shovel nearly as much.
> But outside of Nevada, there are many equivalent distance routes in the US between major population centers, including
And there is nothing stopping those lines from being built other than the lack of will to do it. As before, the will doesn't exist because better options exist.
There is no magic in this world like you seem to want to pretended. All of those things simply boil down to people. Property rights only exist because people say they do, environmental reviews only exist because people say they do, skilled workers are, well, literally people, and the necessary capital is already created. If the capital is being directed to other purposes, it is only because people decided those purposes are more important. All of this can change if the people want it to.
> HN users sometimes have this weird fantasy that with enough political will it's possible to make enormous changes but that's simply not how things operate in a republic with a dual sovereignty system.
Hell, the republic and dual sovereignty system itself only exists because that's what people have decided upon. Believe it or not, it wasn't bestowed upon by some mythical genie in the sky. The people can change it all on a whim if the will is there.
The will isn't there of course, as there is no reason for the will to be there given that there are better options anyway, but if the will was there it'd be done already.
I've taken a Chinese train from Zhengzhou, in central China, to Shenzhen, and it was fantastic. Cheap, smooth, fast, lots of legroom, easy to get on and off or walk around to the dining car. And, there's a thing where boiling hot water is available, so everyone brings instant noodle packs of every variety to eat on the train.
Can't even imagine what the US would be like if we had that kind of thing.
Getting to the airport in most major cities takes an hour, and then there's the whole pre-flight security theatre, and the flights themselves are rarely pleasant. To add insult to injury, in the US it's usually a $50 cab ride to the airport and there are $28 ham-and-cheese sandwiches in the terminal if you get hungry.
In China and Japan the trains are centrally located, getting on takes ten minutes, and the rides are extremely comfortable. If such a thing existed in the US I think it would be extremely popular. Even if it was just SF-LA-Vegas.
Do you mean beginning in the same city? If so, that's downright hilarious. I live 50 miles clear of the city, out in the middle of nowhere, and can be to the airport in the city in less than an hour.
> If such a thing existed in the US I think it would be extremely popular.
I don't know, if people willingly spend an hour getting from one point to another in the same city, the aren't apt to be concerned about how they get from one city to another. I expect they don't put much thought into anything.
Anyway, New York to Las Vegas spans most of the US. There are plenty of routes in the US where rail would make sense. Between Boston, New Haven, New York City, Philadelphia, Baltimore, and Washington, D.C. Which has the Amtrak Acela. Or perhaps Miami to Orlando. Which has a privately funded high speed rail connection called Brightline that runs at 200 km/h who's ridership was triple what had been expected at launch.
I am, thankfully, not.
> Which has a privately funded high speed rail connection called Brightline that runs at 200 km/h
Which proves that when the will is there, it will be done. The only impediment in other places is simply the people not wanting it. If they wanted it, it would already be there.
The US has been here before. It built out a pretty good, even great, passenger rail network a couple of centuries ago when the people wanted it. It eventually died out simply because the people didn't want it anymore.
If they want it again in the future, it will return. But as for the moment...
Since nobody really wants passenger rail in the US, they don't put in the effort to see that it exists (outside of some particular routes where they do want it). In many other countries, people do want board access to passenger rail (because that's all they can afford), so they put in the effort to have it.
~200 years ago the US did want passenger rail, they put in the work to realize it, and it did have a pretty good passenger rail network at the time given the period. But, again, better technology came along, so people stopped maintaining/improving what was there. They could do it again if they wanted to... But they don't.
1. Freight is easier to manage and has better economics on a dedicated network. The US freight network is extremely efficient as others have pointed out. Other networks, e.g., Germany, instead prioritized passenger service. In Germany rail moves a small proportion of freight (19%) compared to trucks. [0] It's really noticeable on the Autobahn and unlike the US where a lot of truck traffic is intermodal loads.
2. The US could have better rail service by investing in passenger networks. Instead we have boondoggles like the California high-speed rail project which has already burned through 10s of billions of dollars with no end in sight. [1] Or the New Jersey Transit system which I had the pleasure to ride on earlier today to Newark Airport. It has pretty good coverage but needs investment.
[0] https://dhl-freight-connections.com/en/trends/global-freight...
[1] https://en.wikipedia.org/wiki/California_High-Speed_Rail
How so?
> The US freight network is extremely efficient as others have pointed out.
'Others' being literally the comment you replied to.
> The US could have better rail service by investing in passenger networks.
Everything there is can be improved, of course, but to what significance here?
https://www.mlex.com/mlex/antitrust/articles/2355294/us-rail...
Boeing tries to compete with Airbus, where the engineers make maybe 70% of what they do in the US, and has upon that made some really bad decisions. Many aircraft companies are profitable.
The problem is the railroads were purchased by the winners. Who turned out to be the existing winners. Who then went on to continue to win.
On the one hand, I guess that's just life here in reality.
On the other, man, reality sucks sometimes.
Imagine if they were bought by losers.
Instead, unless there's a single winner, we will probably see the knowledge on how to train big LLMs and make them perform well diffuse throughout a large pool of AI researchers, with the hardware to train models reasonably close to the SotA becoming more quite accessible.
I think the people who will benefit will be the owners of ordinary but hard-to-dislodge software firms, maybe those that have a hardware component. Maybe firms like Apple, maybe car manufacturers. Pure software firms might end up having AI assisted programmers as competitors instead, pushing margins down.
This is of course pretty speculative, and it's not reality yet, since firms like Cursor etc. have high valuations, but I think this is what you'd get from the probably pressure if it keeps getting better.