The German luxury brands have made the "made in Germany" shtick a core part of their marketing. So Miele, Gaggenau, Vorwerk, ect.
Bosch/Siemens are far larger than those, but they outsourced a lot. But even here, significant parts of the higher-end stuff is still made outside China.
There's no way a modern smart phone or car relies on those ephemeris transmissions. They all just get it from the internet, which takes less than a second. That's one of the reasons why a smart phone has a reliable GPS fix basically instantly after being booted up, while old-school offline GPS units needed minutes to get a fix.
It matters when the level of that body of water drops by a lot in summer and the water temperature rises at the same time. Add environmental laws (cooking the fish is discouraged), and your nuke plant needs to go into safety shutdown pretty reliably every summer.
> The issue is that now the government knows what you are doing online
There's zero technical necessity for this. You could do zero knowledge proofs with crypto key pairs issued together with the eID.
The Swiss proposal for eID includes stuff like that. If a service needs proof of age, you use an app on your phone to generate the response, which is anonymized towards the requester and doesn't need to contact a government server at all.
If we ever end up doing that it would mean terrible things for the state of our regulatory landscape.
A completely optimized high capacity cargo rail line can move 500 rail cars per hour. That's 1000 FEUs if we double stack containers. A lithium battery system in a FEU has around 2 MWh of storage. So that rail line has 2 GW transmission capacity if we saturate it with batteries - the same as a single high voltage transmission line. Being unable to build one of those in parallel to the rail line would be extremely sad.
Note that 500 rail cars per hour is actually an impressive feat of logistics. A normal rail yard at a port would be very happy with a sustained rate of 200 rail cars per hour, and will frequently drop below that.
That also means you can trivially optimize your fuel/cargo ratios. Going across the pacific? Just load 200 more battery containers. Singapore to China right after? Room for 400 FEUs more than normal.
> making the shittiest video with the dumbest script is taking the same amount of energy I'd need to fly across the globe.
I get your overall point, but the hyperbole is probably unhelpful. Flying a human across the globe takes several MWh. That's billions of tokens created (give or take an order of magnitude...).
Does your comparison include training, data center building, GPUs productions, etc or solely inference? (genuine question I don't know the total cost for e.g. Sora2, only inference which AFAIK is significant yet pale in comparison to everything upstream)
No, that's one reason why there's at least an order of magnitude wiggle room there. I just took the first number for J/Token I found on arxiv from 2025. Choosing the exact model and hardware it runs on is also making a large difference (probably larger than your one-time upfront costs, since those are needed only once and spread out across years of inference).
My point is mobility, especially commercial flight, is extremely energy intense and the average westerner will burn much more resources here than on AI use. People get mad at the energy and water use of AI, and they aren't wrong, but right now it really is only a drop in the ocean of energy and water we're wasting anyways.
> right now it really is only a drop in the ocean of energy and water we're wasting anyways.
That's not what I heard. Maybe it was in 2024 but now data centers have their own categories in energy consumption whereas until now it was "others". I think we need to update our collective understanding in terms of actual energy consumed. It was all fun & games until recently and slop was kind of harmless consequence ecologically speaking but from what I can tell in terms of energy, water, etc it is not negligible anymore.
Probably just a matter of perspective. It's a few hundreds of TWh per year in 2025 - that's huge, and it's growing quickly. But again, that's still only a small fraction of a percent of total human primary energy consumption during the same time.
You could say the same about the airplane, does the CO2 emissions that the airline states for my seat include building the plane, the R/D, training the pilot.
Sure and I do, it's LCA https://en.wikipedia.org/wiki/Life-cycle_assessment the problem IMHO being that the AI hype entire ecosystem is literally hiding everything it can about this behind the veil of giving information to competitors. We have CO2eq on model cards but we don't have much datapoints on proprietary models running on Azure cloud or wherever. At best we infer from some research papers that are close enough but we don't know for the most popular models and that's quite problematic. The car industry did everything it could too, e.g. Volkswagen scandal so let's not repeat that.
> hybrid power unit closer to 50/50 split between ICE and electric horsepower
Fun fact, at those ratios it would make a lot of sense to use an electric continuous variable transmission (eCVT) - connect the engine and the motor with a planetary gear set to the wheels, done. The electric motor spins backwards when going slow and forward when going fast. Those eCVTs can be lighter, more efficient an deliver more power across the entire range. But they're illegal in F1 - because they make the car sound boring.
> But they're illegal in F1 - because they make the car sound boring.
I can confirm, my CMAX has an eCVT, and the engine noises are boring. Either it's off, or it's running in a pretty limited range, you can get a bit of fun rev increasing noises if you drive it just right... but mostly boring. My 81 VW Vanagon is much more fun to drive even if it's objectively worse at everything in terms of acceleration, top speed, wheel slip, etc; although the turning circle on the cmax is garbage, so the vanagon wins there. The VW makes fun sounds as you go from low rpm to redline several times as you work through the gears, and the cmax is just droning along.
At least around me, there's a good number for sale. Lots more if you don't mind water cooling and/or subaru engine swaps.
Mostly parts are available; although some things need creativity: federal EGR filters are unobtainable so people might EGR delete instead, rear side markers aren't available, so you have to use Mercedes G-Wagon parts (but they used to use Vanagon parts and scratch off the logos!), I had to adapt a crankcase ventilation valve from a different car because they don't make the ones for mine anymore. Also, headlight switches are available new, but the molds are wrong, so they break when installing (I found a used one, and added headlight relays to reduce the current going through the switch). Oh, and I'm in the second year of ownership, so I started getting fun problems where it runs ok at home but not on the road.
When it is running though, you almost have to drive the snot out of it... Otherwise it'll take an hour to get up to speed.
How Toyota's eCVT transmission works: https://www.youtube.com/watch?v=ppyK3ZlUbtM (Nerd snipe warning: the Weber Auto channel is brilliant and has lessons about all kinds of transmission and engine types).
When it comes to something like F1 I think it's OK for efficiency to not be the top priority. Road vehicles absolutely should be as light and efficient as possible with strict limits on pollution (including noise). But it's OK for society to have a few things like F1 that are just for fun. We just don't want everyone to be driving F1 cars around their neighbourhoods or have an F1 race every week.
I think the F1 teams would all switch to racing versions of those transmissions the second they would be allowed to do so.
The efficiency gains wouldn't even be important in comparison (until you start bringing significantly less fuel than your opponents), but just the reduction in weight and size (important for aero considerations) would be worth it. Also, the power gains from always running the ICE (and its turbo) at the perfect sweet spot in the power curve would be a giant advantage in racing.
Yeah, but a huge part of F1 is the sound. It's iconic in the true sense of the word. I can totally understand them wanting to protect it as part of the brand. Enthusiasts might be into other types of motor racing, but F1 sounds fast. Everyone understands that.
> Well, they could add speakers that make vroom-vroom noises.
And you could have typed this comment into notepad and saved it on a file on your desktop, but instead you shared it with a world that considers it irrelevant.
Definitely, but the weight of the fuel doesn’t matter that much and they allowed quite a bit of fuel. Cars wouldn’t gain much by being twice as efficient if they were any slower.
> And pumped storage is significantly cheaper for seasonal storage than any proposed alternatives.
This is incorrect. There is currently not a single pumped hydro station that is suitable for seasonal storage. They're all designed to drain their upper reservoir in 4-16 hours.
It's the only thing that's half economical. Do the math: Even a modest power plant - 1 GW output - that can run for 1000 hours means you need a 1 TWh (even typing it feels ridiculous) storage reservoir. If you only have 100m of head, that's 3 cubic kilometers of water. That would mean building an artificial lake that immediately would be Norway's 6th largest body of fresh water, and draining it completely every winter.
And effectively, you'd have to build it twice - you also need a lower reservoir. Because there's nowhere to get 3 cubic kilometers of fresh water to fill it otherwise, and you really don't want to do pumped hydro with seawater.
Norway already have seasonal storage with a storage capacity equivalent to 6-8 months of total electricity use in the form of its existing hydroelectric plants, with no need to pump things back up again.
> And yet it's still far cheaper than any other form of seasonal storage.
Only for countries with very suitable landscape, and the willingness to use it - damming high altitude valley is extremely unpopular and bad for the environment.
Also, pumped hydro is expensive. Initial capex is higher than today's lithium batteries, if you design comparable systems. The only reason anybody is still building new pumped storage is that you can use it for 100 years (instead of 20 for the batteries, although nobody really knows how much they'll actually degrade).
I think we'll find hundreds of TWh of seasonal storage elsewhere. Thermal storage is extremely attractive if (or once) you have district heating installed. Takes care of a massive junk of domestic heating, and could take over light (food processing, paper,... ) and medium (chemical, ...) industry. Just don't try to turn it back into electricity...
Once steel and concrete get electrified, we might get seasonal hydrogen storage in underground salt caverns. Concrete and steel need absurd amounts of high heat which probably means making lots of hydrogen, putting those in the right locations might make additional hydrogen for fuel cells/gas turbines available, relatively cheaply (still extremely expensive, seasonal storage always is). But who knows...
Bosch/Siemens are far larger than those, but they outsourced a lot. But even here, significant parts of the higher-end stuff is still made outside China.
reply