[continued from parent]

And the article makes the same prediction that I made above, that if the Trump administration truly wants to restrict international trade on a serious and continued basis, rather than conduct political theater to score domestic points, that’s going to make life a lot harder for competing with China:

Under these conditions, the U.S. will need to accelerate domestic and allied mining efforts while tightening enforcement on chip exports through global cooperation. This will be a challenging task given mounting international resistance to the Trump administration’s potential trade policies

As the United States and China careen toward intensified economic decoupling and geopolitical rivalry, trends in the semiconductor and minerals sectors will define their strategic competition. Both great powers aim to consolidate competitive advantages by hampering the other’s technological development and hammering their trading partners. Both are doing so using increasingly damaging measures—but from opposite ends of tech supply chains. The American position remains strongest in advanced technologies, an edge that the Joe Biden administration sought to preserve and extend through an unprecedented series of export controls. China, meanwhile, is just beginning to implement a parallel export control regime that leverages its dominant market share in critical minerals as well as niche but strategic industries.

Recent tit-for-tat actions mark a troubling new level of severity in this escalating struggle for technological advantage. On December 3, 2024, the People’s Republic of China (PRC) Ministry of Commerce (MOFCOM) imposed its first outright ban on the export of certain “dual-use” critical minerals to the United States. This export control went into force for germanium, gallium, superhard minerals like synthetic diamonds, and imposed additional licensing restrictions on graphite exports. In adopting this ambitious new measure, China was retaliating against U.S. semiconductor chip and manufacturing equipment export controls unveiled only the day prior. On February 4, 2025, in response to new U.S. tariffs on Chinese goods, MOFCOM announced restrictions on additional minerals including tungsten, tellurium, bismuth, indium, and products that include molybdenum. In initiating these outright bans, Beijing has aimed to mirror U.S. long-arm jurisdiction by, likewise, seeking to enforce its export controls extraterritorially in third countries, which could re-export the restricted goods to America.

Juxtaposing “chips” and “rocks” reveals a basic asymmetry between each party’s points of strategic leverage. Beijing is building a dam upstream, threatening to choke off the flow of raw materials and intermediate goods required to produce certain advanced technologies—including semiconductor chips, high-capacity batteries, and a range of defense and aerospace products. Washington’s fortress is further downstream and depends heavily on guarding the intellectual property of American and allied firms employing the technical capabilities of a network of allies and industrial partners. This position has enabled U.S. government efforts to restrict Chinese entities’ access to the latest semiconductors and delay, but not halt, their development of cutting-edge artificial intelligence (AI) capabilities.

My understanding from past reading is that China’s strength mostly isn’t in access to raw materials, but rather in processing of those raw materials. That is, China is not especially unique in terms of what’s in the ground, but rather in that it has large-scale industry to refine those materials, so withholding access to these processed materials permits for leverage. My guess based on past reading as to why processing has gone to China and without looking into individually-processed substances, is that their advantages lie in (1) low labor costs, (2) restricted environmental regulations, and perhaps (3) scale of domestic market and possibly (4) government subsidies.

The first item, high labor costs, is inevitably going to be a US weak point, but we can find a poor-but-friendly country to trade with, probably one poorer than China is in 2025. It’s also possible to possibly partly make use of automation to partially mitigate that; I doubt that this will wholly offset this, though, or manufacturers would have done so.

The second item, restricted environmental regulations, are also probably going to be hard. Maybe some US ones are going to be unnecessary, could be removed, but there are also probably going to be countries that would rather have the economic activity than reduced pollution, so, again, trade is an answer.

The third item, scale of domestic market, is going to be hard to overcome in the longer term. China has a population over four times larger than the US, and even around 2100, after which point the US is projected to have grown and China will have dropped in size, is expected to be about double. China will tend to develop, converge on a per-capita wealth basis with the US. That’s probably going to involve international trade, and not just with one or two countries.

The fourth item, government subsidies, are doable if the US wants to do it, though doing so will weaken other industries. Probably somewhat-easier for the US than China; the US has a larger GDP in 2025.

It’s also important to note that one critical US advantage regarding chip manufacture is in extreme ultraviolet lithography. I understand that this is not something that the US commercialized or presently control, but rather the Dutch, in the form of ASML – the US government paid to develop the basic technology and a prototype, but the Dutch then finished the work to bring it to market. Something that the Trump administration might keep in mind insofar as it is concerned principally with competition with China and not so much with things that Europe cares about, like Russia; actively antagonizing the Netherlands probably isn’t a good idea.

Rare earth elements (REE) have been a focal point in China’s evolving critical minerals policy. As early as 1992, China’s paramount leader, Deng Xiaoping, declared, “the Middle East has oil; China has rare earths.

The US has very little active production, last I looked, but does have inactive production, and Congress was looking at subsidies to remedy that fact.

https://en.wikipedia.org/wiki/Mountain_Pass_Rare_Earth_Mine

As of 2022, work is ongoing to restore processing capabilities for domestic light rare-earth elements (LREEs) and work has been funded by the United States Department of Defense to restore processing capabilities for heavy rare-earth metals (HREEs) to alleviate supply chain risk. [4]

I also seem to vaguely recall that Canada and Australia have rare earth reserves…they just haven’t done extraction, as it hasn’t made financial sense.

https://www.yahoo.com/news/canadas-rare-earth-rush-frontier-200000822.html

As Canada rapidly develops its LNG production and export capabilities and expands its oil industry, the North American country may also be looking to boost its reputation as a rare earth elements producer. Canada has produced rare earth elements (REE) for several decades and is thought to have extensive untapped reserves. It has supported other countries in the development of their REE industries and is now looking to expand its domestic mining activities to help achieve net-zero goals and develop a regional supply chain.

Canada’s 2024 Critical Mineral Strategy Annual Report outlines plans to mine for over 30 critical minerals, with a focus on lithium, graphite, nickel, cobalt, copper, and REE. The U.S. government has repeatedly stated concerns about the growing dependence on China for critical minerals and REE, as well as other energy sources and products, as it looks to develop more regional supply chains. The expansion of Canada’s mining industry could help it provide a stable domestic supply of REE as well as support the development of a North American supply chain. James Edmondson, the research director at IDTechEx, said “It is believed Canada has very large quantities of these materials, even if they have not yet begun processing them in significant quantities.”

That’s maybe a short-term issue, but probably not long-term.

Graphite, essential for lithium-ion batteries, represents a critical vulnerability for the United States, which is in the midst of a $70 billion domestic manufacturing boom in the battery sector

Graphite’s just carbon; basically, very high grade coal. Surely it’s manufactured via refinement of readily-available stuff like coal? The US has, IIRC, something like 40% of known global reserves of anthracite (the next grade down below graphite) within its borders, which is considerably ahead of China, not to mention substantial lower-grade stuff. Also, unless one needs enormous amounts, which I assume is not the case for lithium-ion batteries, any fossil fuel power plant, including crude oil or natural gas, is also going to have as an input something containing carbon, and given energy, that can be reduced to carbon. I cannot imagine that this represents any kind of a long-term constraint for the US.

kagis

This sounds like graphite is indeed obtained by processing coke.

And that is derived from coal.

https://en.wikipedia.org/wiki/Coke_(fuel)

Coke is a grey, hard, and porous coal-based fuel with a high carbon content. It is made by heating coal or petroleum in the absence of air.

Ah, yeah, they mention the Netherlands:

For China, U.S. export controls on advanced semiconductors and chip-making equipment have created a chokepoint for China’s technological ambitions—but with notable examples of innovative workarounds beginning to materialize. These restrictions, reinforced by alliances with Japan, the Netherlands, and South Korea, have sought to block access to key tools like extreme ultraviolet (EUV) lithography machines, which are required to produce high-performance graphic processing units (GPUs), leaving Chinese firms struggling to compete at advanced nodes.

[continued in child]

If the cat is actively trying to push the power button, then that’s probably an issue – he can probably figure out how to flip up a cover and push the button beneath it. But OP says that he’s just accidentally stepping on it. Even if he wants to play with the cover, most-likely not going to actually flip the thing on.

https://en.wiktionary.org/wiki/molly-guard

molly-guard

Etymology: From Molly (female given name) + guard.

Originally a Plexiglas cover improvised for the Big Red Switch on an IBM 4341 mainframe after a programmer’s toddler daughter (named Molly) tripped it twice in one day. Later generalised to covers over stop/reset switches on disk drives and networking equipment.

If you don’t want to make one, can buy premade ones of the appropriate size; for a printer, I assume that you’d just superglue it in place.

https://www.amazon.com/s?k=power+button+cover

I probably wouldn’t normally have looked at a photo gallery of Manhattan signs, but this made it really interesting.

At one point someone explained to me Gorton must have been a routing font, meant to be carved out by a milling machine rather than painted on top or impressed with an inked press.

Every stroke of Gorton is exactly the same thickness (typographers would call such fonts “monoline”).

Monoline fonts are not respected highly, because every type designer will tell you: This is not how you design a font.

Pen plotters need monoline fonts.

I’ve been kind of interested in fountain pen plotters recently, things like these, as I like the look of fountain pen stuff, but would rather use a computer to do stuff (repeatedly, at scale) than train my hand. I don’t think that there’s anything “bad” about monoline fonts. They’re just designed for a specific purpose.

I mean, video games aren’t real either. I played a round of Steel Division 2 earlier today. It was fun, but it didn’t really accomplish anything. The tanks and people there weren’t real – they were just renditions of a computer-rendered world. I don’t think that most people are going to go off on video games as being simply virtual, though.

I wouldn’t personally use the term “boyfriend” or “girlfriend”. They’re fancy chatbots. But I don’t think that there’s anything intrinsically problematic with them. The big issue, from my standpont, is if it causes people to not go out and have kids because the chatbot is taking the place of a partner, exploiting a useful biological imperative – that’s got broader societal effects.

It sounds like in this case, the author is a woman who is a divorcee who was mostly looking for entertainment, not a spouse. So…shrugs

I mean, if she went out and read some romance novels and fantasized, would that be preferable to a chatbot? That’d be more of a traditional route, maybe. But is one clearly worse than the other?

Having a fee of some sort based on payment option is probably desirable. They get charged those fees.

If they don’t expose the fee to you, they still charge the fee, or at least an average that covers their costs across all users, but it gets buried in a single price charged to all users, so you can’t choose whether or not to pay it.

Otherwise, the guys paying via their bank are subsidizing the guys paying via credit cards, even though they aren’t incurring the cost.

every single os

I use sway on Linux. I don’t have shutdown or reboot menus. When I shut down the computer via software, it looks like this:

# shutdown -h now

Rebooting is pretty different, looks like this:

# reboot

I guess that my case has a physical reset and power button that are – while different sizes on mine – near each other. You could put a button wherever you want to, though – those things just short two pins. Go to Mouser Electronics or Digikey or something and get one of those big fancy E-Stop buttons with a protective shield that you have to flip up and run the leads to the reset or power pins, put them on opposite sides of the case, and you can use that. I think I remember reading about someone who used a key switch (like, you have to physically insert a metal key to twist the thing and flip the switch) to power on his computer, just for the aesthetic.

EDIT: Apparently that shield is called a molly-guard and the term was actually originally from computer power switches, prior to making its way to other industrial hardware:

molly-guard

Originally a Plexiglas cover improvised for the Big Red Switch on an IBM 4341 mainframe after a programmer’s toddler daughter (named Molly) tripped it twice in one day. Later generalised to covers over stop/reset switches on disk drives and networking equipment.

but the power usage could easily be 250W+

I mean, a beefy GPU could be ~400W, and a beefy CPU another ~200W. But that’s peak draw from those components, which are designed to drastically reduce power consumption if they aren’t actually under load. You don’t have to power down the components in sleep/hibernation to achieve that – they can already reduce runtime power themselves. One shouldn’t normally have software significantly loading those (especially after a reboot). If you’ve got something that is doing crunching in idle time to that degree, like, I don’t know, SETI@Home or something, then you probably don’t want it shut off.

The reason fans can “spin up” on the CPU and the GPU when they’re under load is because they’re dissipating much more heat, which is because they’re drawing much more power.

• The reason for a clean shutdown is to permit flushing data to the disk. By the time that the OS has gone down, that’s happened. One can just shut off the power once that shutdown has completed.

• The largest issue with unclean shutdowns was that at one point in time, the filesystems used (FAT on Windows, HFS on MacOS, ext2 on Linux) could become corrupt if power was cut at arbitrary times. This has been addressed with later filesystems. Only FAT still sees much use, principally on USB flash drives and SD cards, due to the ability of many operating systems and devices to understand it. While some userspace software can deal poorly with having power cut – I particularly wouldn’t do it while updating an OS – generally shutting down a computer uncleanly isn’t nearly as problematic any more.

• What’s the cost of leaving a computer on, some small amount of power usage? A reboot will log a user out, so there’s no risk of leaving an unattended, logged-in computer. I have my monitor power off, but I don’t normally shut down my desktop when leaving it unattended.

Don’t they support video posts?

kagis

Hmm. Apparently so.

https://help.tumblr.com/knowledge-base/posting-video/

• It’ll need to be a MOV or MP4 file.

• You can post up to 20 videos per day.

• A single video can be up to 10 minutes in length.

• The video upload size limit is 500 MB per video.

• The total video time limit per day is 60 minutes.

Well, it’s not an open platform, but I guess it may be a platform with the resources to serve some serious video, and it’ll be on the Fediverse.

We have been having a lot of discussions about what it would take to get video on the Fediverse at more than PeerTube scale.

I don’t use tumblr, don’t know if they provide a video-centric interface, but I imagine that one could always write a software package to index those videos and link to them. Maybe PeerTube can already do that, haven’t played with it enough to know.

What if it is and it became unrecoverable ages ago?

Yeah, that’s another thing that bugs me about products that can be remotely-updated and especially those which don’t currently represent an ongoing revenue stream. I think that it’s a broader problem, too, not just cars.

I was kind of not enthusiastic when I discovered that TenCent bought the video game Oxygen Not Included and started pushing data-harvesting updates into it via Steam. As things stand, that’s optional. But any company could do the same with other games and not have it be optional. If you figure that all the games out there that have already been sold aren’t actually generating revenue but do represent the option to push and execute code on someone’s computer, they have value to some other company that could purchase them and monetize that.

Then you figure that the same applies to browser extensions.

And apps on phones.

And all those Internet of Things devices that can talk to the network, cameras and microphones and all sorts of stuff.

There’s a lot of room for people to sit down and say “what I have is a hook into someone else’s stuff…now what things might I do to further monetize that? Or who might I sell that hook to who might be interested in doing that?”

Like, if I buy a product, all I can do when I make my purchasing decision is to evaluate the product as it is at purchase time. If the vendor also has the ability and right to change that product whenever they want, then what I’m actually buying is a pretty big question mark. And unless they’ve got some kind of other revenue stream on the line, their only real incentive to avoid doing so is the reputational hit they take…which for failing brands or companies, may not be all that large.

One constraint for efficient markets is that the consumers in it need to be informed as to what they’re buying. If they don’t have that property, you can get market failure. And a consumer can’t be informed about what he’s buying if the person selling them the product can change that product at any point after purchase.

From top-to-bottom:

• HAL 9000 from 2001: A Space Odyssey

• 343 Guilty Spark from Halo: Combat Evolved

• GLaDOS from Portal

• Skynet from The Terminator

• Joshua from WarGames

They are, however, able to inaccurately summarize it in GLaDOS’s voice, which is a strong point in their favor.

to ensure that the technology is “safe, secure and trustworthy.”

None of the really iconic AIs are safe, secure or trustworthy.

That would reduce fuel usage.

I bet that those ad guys haven’t even considered or promoted the fact that they can reduce carbon emissions.

Clearly, the problem is that they went with a pure subscription model instead of also having an ad-supported model. Like, supposing that you’re allowed to turn on the seat heater, but then the car starts playing advertisements while it’s running. They could offer a premium seat heater subscription if you want to buy an ad-free experience.

shakes head sorrowfully

They aren’t very innovative.

https://en.wikipedia.org/wiki/Monty_Python

The group came to prominence for the sketch comedy series Monty Python’s Flying Circus, which aired on the BBC from 1969 to 1974.

No, I believe that that was paid for by the television tax in the UK, rather than interspersed advertisements, as probably most television is.