Avionics can currently only be certified to use GPS (or maybe in theory GLONASS, but I don't think anyone outside of Russia's sphere of influence would do that). In the US, the FAA recognizes a bunch of RTCA guidance documents like DO-208 and DO-229, but those only cover GPS. DO-401 is new (the European equivalent, ED-259, was formally published one revision earlier) and allows use of multiple constellations, but is recognized in the industry as not ready to be certified against. The same is basically true for Europe and the Pacific Rim: they either recognize the RTCA DOs as applicable, or recognize the EUROCAE ED that is harmonized with the RTCA DO.

The jammers on L1/E1 probably affect both GPS and Galileo similarly (Galileo has slightly wider bandwidth on E1, but most of the energy is in the L1 C/A part). Until a year or so ago, most jammers and spoofers were single-frequency and GPS-only -- but new jammers and spoofers are multifrequency and multiconstellation, so even having DFMC avionics wouldn't be a universal fix now.

The long term solution is going to involve beamforming or similar active antenna techniques. Those are also still being standardized, and the Ukraine war is driving the state of the art for military CRPAs.

Ukraine just got a €90 billion loan from the EU. That's enough to cover their financial needs for another 12-18 months. They're starting to sell their designs to Western manufacturers, too, who are intensely interested in cooperation. Ukrainian companies are starting joint ventures outside the country, so that's another cash lifeline, albeit a smaller one.

It is up in the air whether Ukraine gets much territory back. Ukraine is not, however, on the verge of collapse. (Neither is Russia, for that matter, but more later.) Pokrovsk is likely to fall, true, and Ukraine holds only some outskirts, but when I look back to January 1, 2025, Russia was 5km from Pokrovsk, and a year before that, Russia was 35km from the city. Looking at the wider map, the battle lines have not changed that much in two years. Russia has made some advances south of Kupiansk and is contesting Kupiansk itself. If these small cities with pre-war populations of 60,000 and 30,000, respectively, give Russia so much trouble, what is going to happen when they try to take Kramatorsk and Sloviansk, with their pre-war populations of 150,000 and 110,000, if they can even get there?

Russia is not on the verge of collapse, either, but it is on the verge of desperate measures, and may already be in them. It's selling gold internally to take rubles off the market in a bid to both strengthen the currency and provide rubles to redistribute without printing more, with the National Wealth Fund selling off over half its 400 tons of gold to oligarchs. There's still almost 2000 tons of gold left inside Russia, but any significant dip into that is going to cause even more economic problems. A combination of Trump's bumbling and Sunni Arab OPEC states upping their output has resulted in a drop in the price of oil (WTI is trading around $57 today while Brent is around $60). Russia sells at a discount from the market price, and its budget is based on being able to sell oil at $70+ per barrel. It's getting $50ish now, and that may drop even further. Ships are becoming less likely to take on Russian contracts due to the increased activity by the US Navy and attacks by Ukraine. On top of that, while the Ukrainian drone strikes against Russian petrochemical plants isn't decimating the output quite like some think, it is heavily impacting refining capabilities such that Russia put a moratorium on refined products exports, further squeezing their foreign cash income, and they've put limits on internal distribution, which is raising prices, especially outside the major cities. Enlistment bonuses are dropping as local governments run out of money to pay them, with even the relatively wealthy St. Petersburg reducing or delaying bonuses. The federal government is lavishing at least some veterans with education and health benefits as well as some financial benefits, but that appears to not be universal. This is causing a complex situation resulting in high inflation (officially around 7%, probably much higher). The only reason that Russia isn't in hyperinflation right now is tight controls and high core interest rates (16% now, down from 25% because it was absolutely strangling growth). And Russian defense contractors are filing bankruptcy at alarming rates, putting key supply chains at risk.

Despite its population advantage, Russia does have a manpower issue. They're largely not sending conscripts into battle. By law, under the current circumstances, only contract soldiers can be sent to fight (there may be some exceptions, but this is the overwhelming majority of Russian forces fighting). They don't have the money to pay the big bonuses to keep the contract force up, and as veterans return missing limbs and living with severe PTSD, more people are becoming aware of what it's like at the front. Criminals whose sentences were commuted are also returning from the front lines (journalist Emily Hoge has been following this for some time, including the results when hardened criminals become hardened soldiers and return to the streets), pushing up crime rates. Escalating to legal war conditions to allow conscripts to be sent into battle would be an enormous political gamble for Putin.

Russia's problems aren't a lack of physical production capacity (they have that in spades) or people (they have many more they can conscript if necessary). Their problems are that they're running out of money to spend on the war and men they can legally send to fight. Changing the circumstances on the ground requires changing significant political positions inside Russia that have a very strong chance of provoking political backlash, regardless of what levels of control Putin has been able to exert so far.

Maybe Russia can force a ceasefire or even a peace, though it's hard to see what terms would work in such a way that both sides would accept, as both have become skilled at finding something the other side absolutely will not accept, so the war continues. Maybe Ukraine can eventually push the last Russians out of its territory, possibly even out of Crimea, though that, too, is a tall order. Then again, it seemed in the early 1980s that dislodging the USSR from Afghanistan was a tall order. Right now, there's not enough information on either to hazard an educated bet.

Name a specific problem and couch it in terms of memory bandwidth, memory capacity, and FLOPS -- then we can figure out whether GPUs are a good fit. Arm.waving about "data warehouse capabilities" doesn't cut it. Most of the time, the difficulty will be in getting the data you want into the database you want to process it, rather than how far the database can give you an answer.

For example, people don't tend to put databases on GPUs unless they do full scans of the data multiple times each second, because other use cases work just fine with CPU+RAM(+disk). Sometimes that means they have to wait a few minutes instead of seconds for an answer -- but they can do the math for how much that costs them compared to infrastructure. But if they don't have the data imported already, that's gotta of labor and importing it, and GPUs won't help there.

Even expensive GPUs are cheap compared to AI accelerator cards, and the profit margins are lower (in percent) as well. An H200 card reportedly goes for $30,000 and up, about 20 times as much as a RTX 5080. Nvidia's margin (in percent of sales price) is probably two or three times as much for the H200 -- so they need to sell a hugely larger number of GPUs to make the same profit. If AI demand went to near zero, Nvidia would still have a very healthy business but their profits would drop enormously.

Yes, but (a) that number for all of 2024 is less than the 4Q 2025 number for the AVP, and (b) Cybertruck sales in 2025 fell off a cliff compared to 2024, totalling something like 16100 vehicles over the first three quarters.

The actual fucking clods here are the ones who (a) blame unnamed "AI oligarchs" for what randos do and (b) pretend that 501(c)3 means "doing God's Work", especially when the instant example is spamming people with AI slop.

Which "AI oligarchs" are directing the nonprofit that generated these emails? I don't recognize any of the names listed at https://sage-future.org/, but maybe one of them is who you are talking about?

Repeating a lie doesn't make it true. They do look at sales tax, for example in Table 3 -- but primarily that entire paper is a survey of papers that measure tax progressivity, so they talk about how the literature differs. So they don't dwell on sales tax. After all, it represents under 17% of US tax income, compared to 40% for income tax and 24% for payroll/social security taxes. Property tax is 11%, so it doesn't get much attention either.

If you want to be upset about consumption taxes being regressive, look at other OECD countries: they raise much more of their tax revenue from consumption tax, in the form of VAT. The US has much lower rates for sales tax, so the effect of that regressivity is less here.

You're obsessing over roughly a quarter of US rax revenues in order to avoid admitting that the overall tax system is probably the most progressive one in the world. The original complaint in this thread is based on fiction -- a fiction that you apparently want to propagate.

Yes, you are. You ignored the context of that sentence: that Saez and Zucman were being inconsistent in that allocation because otherwise "They argue that the appropriate incidence assumption for distributional
  analysis is to allocate taxes based on remittance obligations." That would mean allocating sales tax to the retailer. The sentence you complained about is not in any way an argument to ignore sales tax, it is merely part of the explanation for why Saez and Zucman were excluded from the literature survey.

Your link doesn't discuss changes in progressivity over time. The US has probably the most progressive tax system in the world (https://www.fraserinstitute.org/studies/measuring-tax-progressivity-high-income-countries-oecd doesn't quantify it for the US as a whole, but "the US tax system is even more progressive than Canada"). It has gotten more progressive over time, including definitely during the Silent Generation and Baby Boomers. See, for example, https://chicagounbound.uchicag... and https://taxfoundation.org/blog...: "The reality is that, since the inception of the modern federal individual income tax in 1913, the US tax code has generally become more progressive, not less." (Emphasis on the original.)

That's absolutist in a rather silly way. A number of statistical "machine learning" techniques are correct by construction -- if you ask the right question. Kernel methods like support vector machines work very well. If they exhibit a "wrong" answer then it means the original problem included an incorrect assumption or omitted important data -- and any technique is liable to give a wrong answer in those cases.

Definitely don't use LLM-generated code without scrutiny, but it's not bad as a starting point or as a source for potential approaches. It's also not bad for code review: I asked gpt-oss to adapt some (not yet tested) code in a certain way, and it noticed a cut-and-paste error in addition to adapting the code. I ended up not using that adaptation, but the big report was helpful.

One can -- and I think should -- be skeptical of lots of things about LLMs, from business models and environmental impacts to quality of output, but one should not disregard them totally.

The name on the Kennedy Center is slightly longer than that:

"The Donald J. Trump and
The John F. Kennedy Memorial Center for the Performing Arts"

Restoring the name just involves removing the first line.

Some of us even use a web site that limits posts to using an eight-bit character set with only text-oriented markup.

Our internal training has shifted entirely to passphrases, to the point that we had to write our own internal training video because every training video we looked at talked about traditional ways of creating a complex password. We found that when people were encouraged to come up with a sentence, they usually came up with something in the range of 25-35 characters, well past the minimums.