that LCOE (Levelized Cost of Energy) is only one side of the equation. Solar has had a low LCOE for several years now, but if you're a business running a solar farm, LCOE isn't what you care about -- profit is what you care about! Profit depends on the cost of solar panels, but it depends just as much on the value of the electricity you are selling to the grid. If the price tends to be near zero (or heaven forbid, negative), profit is dead and the power plant is not worth building.

Wind and solar are the cheapest ways to add capacity to your grid, but they only makes economic sense as long as the instantaneous price of the electricity remains reasonably high. Eight years ago the value proposition looked good because solar wasn't competing too much with other solar, and wind wasn't competing too much with other wind. There are alternative measures that incorporate both cost and value -- see here.

Batteries can restore profitability by timeshifting the electricity sale, but they typically cost well over $100/kWh. They're worth the cost if you can charge and discharge them roughly once a day or more (which can be certainly done if there aren't enough batteries, as in California), but they are not economical for long-term storage and never will be.

We need another solution to reach net zero. For the last 50 years, nuclear could've been that solution, but boy do some people hate it. Nowadays there's another possibility, Enhanced Geothermal Systems, and it looks pretty good. But I still think Molten Salt Reactors are a great design category and we should build them.

Because I'm pretty sure it's actually going to be AI NPCs who generate speech and behavior with GPTs and other neural networks.

Well, UE5's Nanite and Lumen will be big. They don't need raytracing hardware, but I guess it could help.

MAGAs are only like two-thirds of Republicans, so nowhere near the majority of Congress. As Grumpy noted, a majority of Congress wants to halt the Russian advance and Mike Johnson has been stopping them singlehandedly for months (before which, Republicans spent a month unable to agree on a replacement House speaker, since the MAGAs and non-MAGAs couldn't agree and the non-MAGAs wouldn't make a deal with Democrats).

Johnson can't keep his story straight either; previously he made it very clear he wouldn't support Ukraine, but after being elected he pretended he would be happy to support Ukraine but Biden just wouldn't give him a military "strategy" (1, it's the Ukrainians' job to have a strategy and tactics, 2, the strategy and tactics selected depends on available resources which depends a lot on the amount and type of US aid available, 3, wartime strategies and tactics work best if they are not public knowledge, 4, Mike Johnson is lying, as his desire to withhold support hasn't changed.)

just have to pay twice as much for the chips it puts in new FAB 500, FAB 1500, Iskander, and Kinzal missiles it fires at the Ukrainians?

I mean if they want consumer products they can buy finished smartphones and such from SE Asia, so who, if not the military, would be the main buyer of 28nm chips packaged in Russia with a 50% defect rate?

The U.S. stopped sending weapons to Ukraine six months ago (except for some delayed shipments) since all aid packages had expired and the MAGAs blocked any further aid. So it's nice if Russia pays extra for chips, but they're sill dropping something like 10 bombs on Ukraine for every one Ukraine throws back. The bigger ones tend to be GPS guided, which requires chips.

I'm no expert on India, but if you've been paying much attention to India lately, it seems clear that Modi has had enough of this "democracy" thing. Seems he's not only been eyeing Putin, Orban and Ergodan with envy, but taking a lot of notes as well.

A typical discussion of "democracy" in India looks like this:

Yet democracy watchdogs agree that today India resides somewhere in a nether region between full democracy and full autocracy. While democracy-watching organizations categorize democracies differently, they all classify India today as a âoehybrid regimeââ"that is, neither a full democracy nor a full autocracy. And this is new. In 2021, Freedom House dropped Indiaâ(TM)s rating from Free to Partly Free (the only remaining category is Not Free).

So like any "democracy" it needs to limit criticism of the government, publish plenty of books praising the dear leader, cut down on civil liberties, deny that it assassinates political opposition abroad and sure, why not have some official state fact checkers to keep Modi's critics honest?

Let's be clear what we're talking about here. The original CEO, Elena Bunina, stepped down in protest of the war around April 1 2022 and was apparently replaced by Volozh who then also stepped down right after he was sanctioned.

But that was 21 months ago. Which is a big problem if you think sanctions should encourage people to act against the Putin regime. Anyone else who was sanctioned, and who was thinking about whether to stop playing the role or actions for which they were sanctioned, would've looked at Volozh and noticed that he got nothing in return for stepping down!

This is not an isolated case; I've heard of at least one other Russian who was sanctioned, and opposed the war, yet sanctions were not lifted. I can't find his name though, sorry. People need to decide what they want: do you want to defeat Putin in Ukraine, or do you want to merely punish Russians for being Russian?

More broadly I would say the EU has had an unproductive policy toward Russia. For example, Russian anti-war YouTuber NFKRZ has spoken in depth about how Western decisions made it difficult for him to earn money as an anti-Putin YouTuber, to spend his money outside Russia, and to move to the EU. The smart policy would've been: boost the brain drain by granting visas to highly skilled workers who want to move from Russia to the West, offer money and visas to Russian soldiers who hand themselves over to Ukrainian custody, otherwise help Russians move out and spend their Rubles outside Russia, allow YouTubers publishing anti-war videos to a Russian audience to be monetized, and spread pro-democracy messages in Russia. If we're lucky, the West has secret programs to help opposition Russians to oppose Putin inside Russia, but given the lack of smart public policy, I'm not holding my breath for smart covert policy.

"AI is just regurgitating statistical patterns"
"AIs aren't really intelligent or creative"

Sure, let's say that's true. It's also true that a "smart bomb" isn't smart, but can still be dangerous. More importantly, today's GPT technology isn't the goal at all, it's merely a starting point. OpenAI isn't trying to build AI, it's trying to build AGI. Sam Altman & co recently changed OpenAI's mission statement to say "Anything that doesn't help with [building AGI] is out of scope". Several other companies such as Meta, DeepMind, Anthropic and Mistral have similar missions.

"I think AI will probably, like, most likely sort of, lead to the end of the world but in the meantime there will be great companies created" ... "Probably AI will kill us all but until then we're going to turn out a lot of great students." - Sam Altman [source]

OpenAI is very influential. They didn't have to refuse to do the 6-month pause, or change the mission statement to be exclusively about AGI, or ask for $7 trillion, or fire their safety-focused board members. Those were all choices that encouraged the AGI race to speed up. People say "AGI is inevitable, so why bother trying to slow it down?" but... you don't hear that argument about global warming or bioweapons research or genetic engineering. Why are we so quick to give up?

Maybe the issue is that AGI can be extremely beneficial. It could potentially cure cancer, and aging, and homelessness, for example. Humans themselves could do all these things too, but we're slower, and I guess people think it's worth the risk in order to reach utopia a few years sooner. Well, I don't think it's worth the risk and I think OpenAI and Sam Altman are acting recklessly.

AGI is depicted in pop culture as Data / Lore (ST:TNG), C3PO/R2D2, Agent Smith (The Matrix), WALL-E, The Terminator, etc. In fiction, AGIs are almost always basically human in most ways: they are individually unique like humans, they are usually robots or otherwise embodied just as humans are, they typically have humanlike emotions, they are never produced in the billions, they are almost never capable of copying themselves (but thanks Matrix: Reloaded for trying out that idea for once), and they are never much more intelligent than humans (this is a fundamental limitation of human-written stories, but it also allows writers to write happy endings like "humans outsmarted the machines yay!"). People should realize that reality won't be like that.

Normally I would think the same way, but the North Koreans going to the trouble of discovering a new vulnerability and exploiting it suggests that their goal could not be accomplished with ordinary admin rights. This implies that administrator-to-kernel is, empirically, a security boundary against rootkits, even if it officially isn't.

Let's say the Kremlin has hired hundreds or thousands of human "bots" to spread disinformation on the internet (he has, and it's a small line item compared to the rest of the war budget).

Let's even say we can even detect and label 90% of disinformation within 48 hours as state-sponsored (although perhaps only 1% of fake social profiles are caught in reality).

Even then you haven't solved the problem, because the Kremlin is throwing everything against the wall to see what sticks. So in that 48 hours, some American influencer on Twitter, if not Musk himself, will have listened to the disinformation and been convinced by the most claims they find most appealing, and will have repeated them to their 10,000 followers, who immediately spread the word to millions of people. By now it's no longer Russian disinformation, it's American misinformation and cannot be labeled anymore. Right?

If you were expecting sediments (an accumulation of loose particles at the bottom of a body of water) to be impermeable or pinpoint the precise decade or century when certain events occurred, I applaud your optimism, I guess.

Words that the paper uses to describe their findings do not include unusual, remarkable, unexpected, surprising/surprise, unanticipated, unforeseen, astonishing or even startling.

However, they note that "more elongated particles and fibers have reduced mobility" so yeah, maybe date the anthropocene using big immobile elongated particles rather than the microest of the microparticles.

People don't think they will get it all back, they think they could. Billions of dollars of investments have been known to sometimes go up even more than the cost of Super Bowl ads, lawyers, campaign contributions, charitable donations and ... sorry did you say toilet?

Hopefully everybody can finally get behind an "all of the above" energy strategy now, starring not just the usual Chinese solar panels and Chinese/EU wind turbines, but also North American Enhanced Geothermal Systems, French nuclear plants, UK/US/Canadian Molten Salt Reactors and... say, did you ever notice we do emissions targets all wrong?

The goal should be to reduce emissions globally, not just in our own country. From that perspective, who is the global leader in reducing emissions? China produces more than twice the carbon emissions of the United States (though China produces less per capita), but let me ask you a question: why are we building so many solar farms now? Because of subsidies? They're affordable without subsidies now! Many countries have been building without subsidies for years now! But who made them affordable? China! So after you after you consider China's contribution to actually solving the global warming problem, aren't they doing pretty well actually? (And sure, those old subsidies that were originally needed to afford solar panels helped pay for factories producing cheap panels on a huge scale. Even so, a lot of voters would prefer money be spent on local factories over Chinese ones.)

And the funny thing is, looking at it this way is way easier politically, yet Greens don't do it. I remember seeing The Green Party of Canada's platform and something like 80% of their climate change plan was about punishing and then banning fossil fuels, with no mention at all of nuclear or enhanced geothermal, let alone manufacturing anything in Canada. So how popular is the Green Party in the oil-rich prairie provinces? Well, they got 2.3% of the vote nationwide.

Imagine if instead their strategy were manufacturing and exporting clean Enhanced Geothermal technology or Small Modular Reactors. Wouldn't people actually vote for "retraining of oil workers for geothermal drilling! wind power jobs! nuclear power jobs! heat pump manufacturing jobs!"? Just a thought. (as for solar jobs, Canada needs much more energy in the wintertime, and are you familiar with that white, cold stuff that falls on top of solar panels from the sunless February sky?)

Regulations and NIMBYs have been a big problem too. As the CEO of Fervo Geothermal noted, "we don't get the same tax credits as the... other renewable industries get"... "on the permitting side, the energy act of 2005 put in...for oil and gas drilling on federal lands...more streamlined development...but that policy did not extend to geothermal, so geothermal has to go through a much longer and much more intensive permitting process than oil and gas wells do." Luckily "The inflation reduction act...for the most part...we've finally be put on a level playing field...from a tax credit standpoint." Okay, but what about geothermal permitting? And oh God all the horror stories I've heard about nuclear power regulations, not to mention NIMBYs...

This is incorrect. It's true that there is "committed warming", meaning that positive feedbacks caused by past warming will lead to future warming. However, this can be roughly balanced by an opposite effect: natural carbon sinks (especially the oceans) that reduce CO2 concentrations.

As this article by Carbon Brief puts it:

Much of the confusion around committed warming stems from mixing up two different concepts: a world where CO2 levels in the atmosphere remain at current levels; and a world where emissions reach net-zero and concentrations begin to fall.

I also think a lot of this desire to ban plastic bags nowadays is guilt-by-association.

• "Oh isn't the ocean just filled with plastic now? Yeah, we'd better ban it here in our city a thousand miles away from the ocean." Okay, I would just like to point out that two-thirds of rivers do not contribute to the problem so there must be some way of keeping bags out of rivers without banning them completely.
• "Plastic is made of petroleum! Oh no, isn't that what causes global warming? I guess we should ban it then." Uh, greenhouse gases caused by burning petroleum causes global warming. Plastic itself is sequestered carbon. Besides, isn't the amount of petroleum used to make plastic very small compared to the amount of petroleum that we simply burn? And isn't the amount of plastic used in shopping bags just a small fraction of that?

We wouldn't be "literally" choking the planet even if it were possible.

The reason recycling became popular was twofold:

• 1. Plastic is not biodegradeable.
• 2. In the 1980s or so, people thought we were quickly running out of landfall space.

Regarding (1), that's changing. Since plastic is made of organic molecules that can burn in fire, chemists and biologists might well have predicted this decades ago: microorganisms can theoretically eat it as food, they just hadn't figured out how. But then evololution happened.

Regarding (2) I haven't heard of landfill space being a problem for decades. So, we could simply stockpile plastic instead of recycling it. Either recycling technology will make improve, making it much cheaper in the future, or we'll eventually run low on the petroleum typically used to make plastic (so that recycling becomes cheaper than making new plastic), or we can just wait until plastic-eating bacteria are everywhere, then put the plastic outside and wait for it to vanish..