Things have cost floors. You can't profitably make something below this floor, because even as the cost of your value-add approaches zero, the inputs still cost something: materials, electricity, etc. Lithium batteries, due to the volume of manufacturing and global competition, are approaching this point. However, despite the drop in prices over time, you can see it's leveling off in recent years. This is why. The same pattern is apparent is solar panels, displays, mobile phones, etc. All the things that are made in huge volume by aggressive competitors are driven towards some cost floor.

The batteries are made of materials approaching the cost of dirt. Like everything, sodium batteries also have some cost floor. But the floor is lower.

Right now, sodium is not substantially lower cost. That's because the volume is a small fraction of Lithium cell manufacturing: lower volume and less competition. As volume grows, and it will, because there are objective benefits for real applications, the cost will fall, and ultimately be lower than Lithium.

This is a win for everyone. All the Sodium battery hate seen here is deeply stupid. You'd think this place is full of Chinese Lithium battery manufacturers, which is about the only conceivable rationale I can imagine for this behavior.

"believe clean energy around nuclear is going to be the answer"...

Yeah, that''s what the AIs in The Matrix said just before turning us into pod people.

The sun is the only one clean energy source - the rest are dirty to varying degrees. Nuclear is definitely not clean.

Someone well qualified in nuclear fuel chemistry that I know doesnt think SMRs are going to take off. He says people have been talking about them for decades.

SMR will not happen in the US. Nuscale's UAMPS project in Utah is an case study in why such hopes are futile. Costs escalate as regulators embiggen requirements around new designs, and pressure groups spawn like weeds to delay everything until the financing collapses.

What would it take? Two things: 1.) Deep pockets that can fully finance the project without bonds and shifting interest rates, all up front. 2.) Sincere and unwavering legislative support, including legal exceptions from state and federal regulatory sabotage and pressure group legal attacks.

The former is conceivable, but we're talking about someone committing $10B+ for ten or more years, and that's hard for anyone, even Big Tech monsters: Yes, there are trillion dollar companies, but like all such oversimplifications, they aren't so liquid they can easily just sink that kind of liquidity into a nuclear project. That's why they're grifting around, trying to wheedle deals with state government and regional utilities.

The latter is fiction: executive orders and other half measures are not sufficient. What would be sufficient is a bill, passed by the Federal government, signed by a president, capable of surviving challenges in front of SCOTUS, because it will go that far. That model has happened in the US in the past: TVA etc. Today, our leaders are balkanized around extreme pressure groups and vocal minorities, and no rationality is apparent any longer. The US isn't a place where things like this are feasible today.

You're conflating non-deterministic with chaotic. You can have a perfectly deterministic system that exhibits chaotic behavior, meaning that the tiniest difference in input often produces a wildly different output.

I think hallucinations may just turn out to be an essential feature of sufficiently-rich models. People hallucinate, too. Mostly we're better at identifying our hallucinations. Mostly.

I think cryptocurrencies are super easy to understand, and the last clause of your sentence just spelled out everything anyone needs to know.

If you're into cryptography, you might find the details of how they work interesting (well, moderately; there isn't any really interesting cryptography there), but none of that has any bearing on whether it makes sense to put your money into it.

Nonsense.

It's debatable whether HFT is a net positive for the market (it does provide positive value, ensuring that pricing is consistent across markets), but it definitely doesn't "rig" the market. If you're trying to compete with the high frequency traders at their game, you're going to lose, but you're also doing it wrong. For people who don't spend all their time reading quarterly reports, the right way to do it is to put your money in a decent mutual fund or index fund, and let it grow. If you want to get more aggressive than that, then the way to do it is Buffet-style value investing: Identify companies with good products, in good markets, with good management and good long-term strategy, and buy and hold.

It's got more than that behind it, too. It's backed by a massive network of debt contracts obligating people to do work in order to generate value in the form of dollars. That's how dollars are created: When banks issue loans they create the money[*] they're lending out at the same instant they create the loan contract that obligates someone to repay it.

Fiat currencies (which BTC is not) are essentially built on the legally-enforceable promise of people to do work in the future. They're backed by human energy and effort.

There are some science fiction stories in which civilizations have found a way to store large amounts of energy safely in small coins, which are used as currency. In those fictional civilizations, when you give someone money you are literally giving them power that they can use to do things. Some fantasy worlds do the same with congealed magic as currency (I recently read the "He Who Fights with Monsters" series, which has a world that does this). The value of such a coin is what you can do with it, completely aside from anyone's opinion about the value, and what you can do with it is defined by the contained energy. That would be the perfect currency, because the coins' worth is 100% intrinsic.

But fiat currencies are almost as good! They aren't encapsulated power to do stuff, but they represent a promise by some person or people to do stuff. They're very nearly 100% intrinsic value (unlike, say, gold, which has a little intrinsic value plus a lot of speculative value).

Bitcoin, on the other hand, is 100% speculative, with nothing at all backing it.

[*] It used to be the case that they arguably only created ~90% of the money they lent, but the reserve requirement was reduced to zero. Even before then it was debatable because while loans "tied up" reserve capital, requiring the banks to hold it in reserve, they did hold it, they didn't lend it out. In any case, with a 0% reserve requirement banks definitely now create every dollar they lend.

If it comes to that, the difficulty can be turned down to make mining profitable again.

I wish it would just end.

You're completely wrong to equate bitcoin with the stock market. When you buy shares in a publicly-traded company you're buying something real, a portion of an actual business that owns real stuff, has real employees, makes real stuff and sells it to real people. Exactly how much that company is worth, so exactly how much your share is worth, isn't really clear because calculating that value requires a lot of guesses about what the company, and its market, and the economy as a whole, are going to do in the future. But it's definitely worth *something*.

That's not true of bitcoin. You won't hear people talking about bitcoin "fundamentals" the way they do with stocks, because bitcoin doesn't have any fundamentals. It's just vibes, all the way down. Oh you can talk about hashrates and work factors and the cost of mining equipment, etc., but none of that has any actual value. If the governments of the world united and successfully banned cryptocurrencies, all those ASIC rigs would be useless. If they were still using GPUs at least you could repurpose them for AI, or graphics.

It's amusing watching people cling desperately to the old "Climate Change" Narrative when it's been thrown under the bus because we need massive, reliable power generation to keep AI data centres running.

The 'experts' got it all wrong again.

One of our friends does infrastructure work all over Canada. He has a big tank of diesel in the bed of his truck so he can refuel when he can't get fuel locally and the next gas station is 200 miles away.

EVs will have to carry generators I guess.

> which is a thing that isn't really happening.

The Federal government literally just paid this city to change zoning rules so developers can knock down single-family homes and build blocks of condos with zero concern about whether the infrastructure can support it.

But, yeah, it's "not really happening".

So your solution is to make the roads worse so people can't drive as far?

The people I know with expensive houses in Toronto never plan to pay off the mortgage because they expect to be able to sell the house for several times what they paid for it, take the profits and buy a house for cash in a cheaper province when they retire. Or move to somewhere like Mexico.

If they're typical then the Toronto housing market is just like renting except people expect to get a big bonus at the end of the rental. And based on other stories I've heard it seems to be increasingly propped up by fraud.