Coal demand is elastic. As the price of coal rises or falls, demand decreases or increases. This isn't true of all commodities, but it's true of all commodities that have alternatives. In this case, coal's primary use is power generation and there are many alternatives. A crucial alternative is natural gas, which produces far less CO2 than coal per kWh generated, and is also a better solution than coal for addressing the intermittency of PV and wind, which are the cheapest and lowest-emitting alternatives. Which is why at the same time the Biden administration is moving to reduce the supply of (and therefore demand for) coal, it is moving to increase the supply of natural gas.

This is a good and well thought-out strategy to reduce CO2 emissions at relatively little economic cost.

It's also a politically-savvy move. There is a very small but fairly loud minority in the eastern US that depends on coal mining for their livelihoods. Shutting down that coal production would be politically disastrous. Preventing the expansion of coal mining in the western US will actually help to prop the price of coal up. Propping up the price of coal will reduce demand for coal, but in the short term it will help the existing mining operations to remain profitable. Eventually those eastern mines will close, too, either because demand falls too far or because they are played out, but that's a problem for a future administration.

Personally, I think a carbon tax would be a better strategy, one that would require less direct government meddling with supply and demand and would harness the power of free markets to most rapidly and efficiently reduce emissions, at minimal cost. A carbon tax would also allow us to eliminate subsidies on solar and wind, and would likely make nuclear more competitive. But a carbon tax (and corresponding carbon tariffs, to prevent us from just pushing the emissions overseas) is politically intractable at present, so the government has to use these sorts of narrower and less efficient methods.

If you look at economics that throws out the external costs of coal.

Globally fossil fuels recieve seven trillion dollars annually in public subsidies. But that's just a drop in the bucket compared to the costs it is allowed to pawn off on other parties. If fossil fuel users had to pay the externalized cost of pollution, then the world would be running on nuclear power right now.

I don't think Econ 101 price/quantity equilibrium is entirely what's going on here. Gigabit service *availability* is about the same in Spain and the US, despite America's per capita purchasing power adjusted GDP being about 60% higher than Spain's.

I think the relevant figure is this: Spain has roughly 2.8x the population density of the US. It's surely a lot more expensive to build the infrastructure to cover roughly the same percent of the population here.

Although that's not *nothing*, the question is who exercises admnistrative control of that data. If the Chinese government demands data from ByteDance's management, and ByteDance's management complies, that data is not safe. Of course, even in the US a federal agency can obtain a secret warrant which enables them to help themselves to your private data held by a third party, and because it's *secret* you can't challenge the warrant's legality.

The smart thing is don't put anything sensitive onto any kind of social media. Now some metadata may itself by sensitive for certain persons, like your approximate location at various times. Such persons shouldn't use social media at all, even if the data is hosted in the EU, which generally has the best data privacy protections in the world, because there is *no* country in the world where a company can defy a lawful warrant, whatever "lawful" means in that country.

Nah, you'll get even more stupidity. Trump is proposing blanket tariffs on everything from China.

If that were the only effect, it would be a bad deal. But the track record of the CHIPS Act investments so far is really good. In just a couple of years, the act (with some help from the IRA) has nearly tripled US investment in manufacturing, with most of the growth focused on semiconductors and batteries. The federal government has provided only a few hundred million in seed money in targeted areas (like this one), but that has produced $150B in new construction of manufacturing facilities. Most new manufacturing employs relatively few people directly (like the 160 in this case), because US manufacturing is heavily automated. But it indirectly employs a lot more, especially in companies that build and service manufacturing robots, and all of their suppliers, etc., on down the chain.

I'm a free market libertarian and I'd generally prefer to see government stay out of stuff like this, but it really looks like these federal programs are incredibly successful, with relatively small taxpayer investments generating 3 orders of magnitude larger private investments, all in high-end manufacturing that will likely generate returns that are another 1-2 orders of magnitude larger than that. If a federal dollar can generate $10,000 in new economic productivity, that's a big win, including for taxpayers since that $10k will be taxed, returning far more than was spent.

Of course, it remains to be seen whether the new manufacturing construction will actually generate the returns. Maybe it'll turn out that US chipmakers can't compete with Asian ones, and they'll all just fold. The fact that the vast bulk of the investment is from private sources, though, indicates that there are people who strongly believe the US can compete.

I don't really understand why the federal money has this effect, though. The large private investments seem to indicate that people think these are good bets, but why didn't they think that before the Biden admin started seeding them? The actual federal cash is such a small percentage that it can't be what's motivating the investments. Perhaps the threats of increased tariffs on Chinese chips and batteries are most of the story? But Trump was doing that six years ago. I don't understand, but won't argue with success.

some of us use

Ctrl-Ins and Shift-Ins

I think one place to expect operational savings is refueling. Conventional reactors spend about 8% of the time offline being refueled. Every eighteen months thousands of workers from all around the country come to the site to do the work. SMRs are designed to need refueling much less often, typical every 3-7 years. Some designs go for up to thirty years without refueling. Plants with a larger number of smaller reactors can also do maintenance and refueling without losing any revenue, as the remaining reactors put out a little more power to compensate.

SMRs shut down and cool down much faster; some don't require any active cooling measures at all. You just shut the thing down and a week later it's cold even if you don't have outside power. So there's a lot less plumbing to monitor and maintain.

Of course these are all just promises now. Running these things is going to be so different we won't really know until we've built and operated some.

"If we only did the right thing then we couldn't do what we want to do."

Which in itself says nothing whether you are or are not violating the creators' rights.

You as the non-owner of the IP have certain fair use rights that depend, not on the mechanism by which you obtain a copy of the data, but on the effect of what you are doing with the data upon the copyright holder's proprietary interests. A download button does *not* indicate content is free game for commercial use.

It's already extremely safe. Cycling has a lower death rate per participant than *tennis*. And while your risk per *mile* is signifiantly higher on a bike than as a passenger in a car, your risk per *hour* is signifiantly lower. Since most cyclists aren't putting nearly as many miles on their bike per week as their car, the bike represents a low risk to them; in fact if you take up cycling your chance of dying in the next year goes down by 1/3 once the fitness benefits kick in, even though you've just added a new way to die to your personal list.

So tech like this is unlikely to reduce *absolute* risk very much, because absolute risk is already very low. It so happens this particular tech could reduce the most fatal type of accident -- being struck by an overtaking motorists -- but these types of accidents are very rare, as are cycling fatalities. Since there's only about eight hundred cyclist mortalities / year in the US there's not a lot of room for improvement, especially as this tech is bound to be installed on only a tiny minority of bikes. It does nothing for the two most common types of accidents: (1) cars entering the street to make a turn and hitting a cyclist traveling along that street and (2) cars passing a cyclist and making a right turn at an intersection across the cyclist's path (the "right hook"), so it's unlikely to affect metrics like ER visits and hospitalizations very much.

We have to sharpen our thinking about what we're actually trying to accomplish when we talk about "making cyling safer". I'd suggest there's two things we can be reasonably trying to do: eliminate as many *preventable* deaths and injuries as possible and make people *feel* safer when riding a bike. There's a lot of injuries that can be taken off the table by designing and marking intersetions better and improving lines of sight. Many of these changes would also reduce car-pedestrian accidents and car-car accidents too.

Technologies like this can't make cycling statistically much safer than it aready is. But they can do a lot to make cyclists feel safer -- much the way some cyclists are spending hundreds of dollars *today* on rear-facing radar units. Those are good things, but they're no substitute for better design which would both make cyclists feel safer and make everyone statistically safer.

Sure, if you just fixed people, that would work better than any conceivable technical solution.

But you can't fix people, so there's no point in complaining that *some* people are *sometimes* stupid, careless, or irresponsible..That will never change. Sometimes, even, that careless person might be *you* on a bad day. We all rate ourselves based on our performance on our good days; which is why we all think we're better-than-average drivers, but really our risk to ourselves and others is dominated by the days we didn't get enough sleep, are stressed out, distracted, and running late. Those are the days when the things we habitually do right go out the window.

If you frame the root of the problem as being "people aren't good enough to operate the world we've built", then you're stuck putting band-aids on the problem. That's not unreasonable as it sounds, think of how things would be if you didn't *have* band-aids. People would be getting limbs amputated because otherwise insignificant cuts got infected.

So it's perhaps better to frame the problem this way: the world we've built is not suitable for people to operate in with acceptable safety. The root of the problem is design, so that should be highest priority. But even so, we'll still need to address the failures that better road design can't fix. Even the most safely run factory still needs a first aid kit, and that first aid kit is going to be stocked with band-aids.

But it matters to the research and operation teams whether their downtime is two months or nine months. Such a reduction could alter the economics of a robotic exploration program, which would surely be a prelude to a manned mission. So the robotic program could both provide input into planning of the manned mission while proving the propulsion technology is reliable enough to be man-rated.

Current concepts for a manned Mars missions would last 22 months, 21 of which would be spent in transit and one on Mars. If the round trip transit time were reduced to 4 months, you could spend 18 months on the surface in a mission of the same length. In such as scenario with robotic missions you could avoid staging supplies that only *might* be needed knowing you could send them if the need arose *during* the mission. You could respond to unexpected circumstances, or return samples to Earth for analysis that could alter the priorities of the mission.

So it could make sense to build unmanned vehicles with such a technology -- as part of a manned *program*.

Considering they're stackable, why do you think you need to do them all at once? Would one be better than none? Do you know, by measurement, how much power you actually consume at night, when the panels aren't generating?

Not criticizing, just curious and gathering info.

You're confusing V2L (Vehicle to Load) with V2H (Vehicle to Home) or V2G (Vehicle to Grid). V2L means you can plug normal 120V/240V AC stuff into your vehicle and it'll run. I have an Ioniq 6 that can do this and it can provide up to 3.6 kW of power, with 3.6kVA for 230V power devices and 1.9kVA for 120V power devices. It has a maximum draw of 15A. You aren't running your house on that, but can do some emergency loads.

V2H is really only available with the Ford F-150 Lightning right now. I *think* you can make it work with a Nissan Leaf because CHAdeMO has supported V2H from the beginning, but it is uncommon because of the stress it puts on the Leaf's non-temperature regulated batteries.