windy.com with a VPN in a private browser window. They can’t track you if they don’t know where you are!

Yeah, I concede that small caps are more likely to be carried away by rainwater than whole bottles :D. What I meant was that for every loose cap on the ground there is a bottle lying around somewhere, and also there are bottles with caps on. No one is tossing their cap into the bushes and then taking the bottle to the recycling center.

I pick up street litter, and having picked up thousands of pounds, I have never felt that loose caps are a problem, let alone one that requires such a solution. The number of littered bottles, with or without a cap, is greater than the number of loose caps, and the amount of plastic in every bottle dwarfs the plastic in a cap. Fixing the cap to the bottle will do nothing to improve the recycling rate of plastic if entire bottles are already tossed anyway.

I consider the idea of cap tethers as adversarial memetic warfare thrust upon us for some unknown ulterior purpose, possibly to make us hate the very idea of environmental consciousness. Same as paper straws. I like plastic bag bans though.

As far as picking litter is concerned, I personally prefer finding bottles without a cap. At least those are empty, all liquid having evaporated after the bottle has spent several months in the bushes. The capped bottles are often half-full and are just nasty. (Who even pays for a bottle of drink and not drinks half of it anyway?)

It is important to remember that, unless accompanied by convincing evidence for selective advantage, any single inheritable trait is more likely to have arisen from genetic drift, not from natural selection! There is, in my opinion, too much focus on conversation about superficial phenotypic traits like “shape of the nose” this and “angle of the eye” that, all the arguments about how one is better than another. Could the asiatic epicanthic fold give advantage against icy winds? Maaaybe… But it doesn’t even have to. What about the asiatic dry earwax gene? You’d struggle to even come up with a story of how dry earwax or wet earwax is actually better under certain conditions, or you could just say “it’s a single nucleotide polymorphism that could have spread by genetic drift” and be done.

Very few human traits have definitely been naturally selected for: light skin in non-sunny climates for better vitamin D production, sickle cell gene for malaria resistance, lactase persistence for animal milk consumption. Even there, the estimated selective advantage is actually much smaller than you’d expect: lactose tolerance confers only something like 1% advantage! There are many more possible neutral mutations than advantageous ones, and each one has a chance to be fixed in the entire population by genetic drift, meaning that any widespread human trait that is less clearly advantageous than lactose tolerance is more likely to be neutral than advantageous at all.

Even mildly disadvantageous mutations can be fixed by genetic drift, especially in humans since we have had many bottlenecks and founder effects. There was an area in Appalachia populated by blue-skinned people due to founder effect. No one is going to try to argue how having blue skin was actually advantageous for them to blend into their environment! There is an area in Dominican Republic with a very high rate of children born intersex, again due to a founder effect mutation. They are not considered exceptional and live normal lives as their culture has adapted to treat them as routine, as a kind of third gender. But they are not some kind of new level of human evolution, an adaptation for an intersectional era!

The only mutations that definitely cannot spread by genetic drift are those that definitely kill you.

This is a great gotcha! I just recently learned that 700C, the “29er”, and (some) “28 inch” are all the same wheel.

To be fair, when I was little I too was guessing that “C” stands for centimeters or something metric. Now I know that “C” in “700C”, the most popular road/hybrid wheel size, stands for the third size in the French “ABCD” notation, where sizes “700A”, “700B”, and “700D” are obsolete and are no longer manufactured.

I’d love to use ISO sizes, but even if I know that I need a 40-622 wheel, there is no way to search for it on the storefront if every single seller made gross mistakes in labeling their product! I have to ignore the specs shown entirely and make educated guesses based on title alone. For example “WHEEL AL 700 FRONT ALEX AP18 QR Silver UCP” in the picture is almost certainly a 700C wheel and NOT an 18-inch wheel. The “18” in the title probably stands for 18mm rim width, which means that this wheel will fit my bike and tire, but is a bit more narrow than ideal 23mm. The sellers must be copying the title verbatim from the manufacturer, and then haphazardly filling out the specifications without knowing or understanding the actual numbers. The ISO size is not mentioned at all.

Given a radiative forcing coefficient of ln(new ppm/old ppm)/ln(2)*3.7 W/m**2 I have previously calculated that for every 1kWh of electricity generated from natural gas, an additional 2.2 kWh of heat is dumped into the atmosphere due to greenhouse effect in every year thereafter (for at least 1000 years that the resulting carbon dioxide remains in the air). So while the initial numbers are similar, you have to remember that the heat you generate is a one-time release (that dissipates into space as infrared radiation), but the greenhouse effect remains around in perpetuity, accumulating from year to year. If you are consuming 1kW of fossil electricity on average, after 100 years you are still only generating 1.67kW of heat (1kW from your devices and .67kW from 60% efficient power plant), but you also get an extra 220kW of heat from accumulated greenhouse gas.

I have wondered this question myself, and it does appear that the heat from the fossil/nuclear power itself is negligible over long term compared to the greenhouse effect. At least until you reach a Kardashev type I civilization level and have so many nuclear/fusion reactors that they noticeably raise the global temperature and necessitate special radiators.

I wanted to go with this, but had to go even bigger. The largest mine truck according to the wiki is BelAZ 75710 (as seen in the picture) which has “59/80R63” tires that are merely 402.7cm big.

Physical players disconnected from the internet can still receive offline firmware updates included on the discs themselves. The moment you insert a new disc, it automatically executes BD+ code that in theory could patch the firmware to blacklist an arbitrary old disc that you own. This has never yet happened with a previously-legal disc, but then again for example Amazon has never deleted purchased copies of the 1984 book from customers’ kindles, until one day when it did.

a standalone drive

Another cool/scary feature of the BluRay spec is offline firmware updates (called BD+). Any disc can contain code that runs automatically and can patch the player firmware or execute arbitrary functions. So if you have an older hacked player and you insert a newer disc into it, the AACS Consortium has the ability to brick it. Or if you “own” an older disc but the Consortium starts to dislike it for some reason (maybe they discovered that the disc was printed by a pirate publisher, or maybe there was a retroactive licensing dispute), they can include code on every newly published disc that blacklists the old disc. Even with a standalone player that you never connect to the internet, the moment you insert any new disc into it, your old “problematic” disc will be unplayable. This has never yet happened with a previously-legal disc AFAIK, but it is possible within the spec. Every player manufacturer must obey the spec and implement the BD+ virtual machine in order to be allowed to read AACS content. And if you hack your player to ignore BD+ code, then the newer disc will not play because its content may be scrambled in a way that only the custom BD+ code included with it can unscramble.

But then you have to watch Picard 😢

A funny culprit I found during my own investigation was the GFCI bathroom outlet, which draws an impressive 4W. The status light + whatever the trickle current it uses to do its function thus dwarfs the standby power of any other electronic device.

That’s how I found out that my desktop speakers consume power even with the physical button being off and status light dark. The power brick stays warm indefinitely, a good 20W feels like! I have to unplug that thing now when not in use. Any normal power brick will be <1W of course.

FYI, the magic about:config key that you need to set to false is “keyword.enabled”. After that Firefox will finally stop using any non-url string as a search query and will instead say say “Hmm. That address doesn’t look right. Please check that the URL is correct and try again.”

I am still sad Hitachi was too embarrassed to carry on the legacy of its name and sold off the Magic Wand brand to its subsidiary manufacturer. Hitachi, the brand name was a compliment to you, not a liability! You lost out.

Some notes for my use. As I understand it, there are 3 layers of “AI” involved:

The 1st is a “transformer”, a type of neural network invented in 2017, which led to the greatly successful “generative pre-trained transformers” of recent years like GPT-4 and ChatGPT. The one used here is a toy model, with only a single hidden layer (“MLP” = “multilayer perceptron”) of 512 nodes (also referred to as “neurons” or “dimensionality”). The model is trained on the dataset called “Pile”, a collection of 886GB text from all kinds of sources. The dataset is “tokenized” (pre-processed) into 100 billion tokens by converting words or word fragments into numbers for easier calculation. You can see an example of what the text data looks like here. The transformer learns from this data.

In the paper, the researchers do cajole the transformer into generating text to help understand its workings. I am not quite sure yet whether every transformer is automatically a generator, like ChatGPT, or whether it needs something extra done to it. I would have enjoyed to see more sample text that the toy model can generate! It looks surprisingly capable despite only having 512 nodes in the hidden layer. There is probably a way to download the model and execute it locally. Would it have been possible to add the generative model as a javascript toy to supplement the visualizer?

The main transformer they use is “model A”, and they also trained a twin transformer “model B” using same text but a different random initialization number, to see whether they would develop equivalent semantic features (they did).

The 2nd AI is an “autoencoder”, a different type of neural network which is good at converting data fed to it into a “more efficient representation”, like a lossy compressor/zip archiver, or maybe in this case a “decompressor” would be a more apt term. Encoding is also called “changing the dimensionality” of the data. The researchers trained/tuned the 2nd AI to decompose the AI models of the 1st kind into a number of semantic features in a way which both captures a good chunk of the model’s information content and also keeps the features sensible to humans. The target number of features is tunable anywhere from 512 (1-to-1) to 131072 (1-to-256). The number they found most useful in this case was 4096.

The 3rd AI is a “large language model” nicknamed Claude, similar to GPT-4, that they have developed for their own use at the Anthropic company. They’ve told it to annotate and interpret the features found by the 2nd AI. They had one researcher slowly annotate 412 features manually to compare. Claude did as well or better than the human, so they let it finish all the rest on its own. These are the descriptions the visualization shows in OP link.

Pretty cool how they use one AI to disassemble another AI and then use a 3rd AI to describe it in human terms!

Can’t access the article, but wasn’t China the one most vulnerable from the Malacca Strait being a chokepoint? As in, their trade towards Europe and fuel from the Middle East being potentially threatened? How does Thailand pitching to the US make sense then? How would a Thai bypass even increase security, since both routes are in the same area and can be equally blockaded? There aren’t any problems with throughput capacity at Malacca, unlike say at the Panama Canal. Maybe it will make the travel distance slightly shorter, but is there really any way it could ever be cost-effective to offload and reload ships for a few hundred kilometers savings?

Thank you for your detailed input!

It’s not even a platonic ideal - it’s drawing a supply/demand curve and thinking you understand how prices work in a market economy.

You got me 😁. I love drawing supply-and-demand curves. Seems pretty hopeless then if to even begin to understand how to vote “correctly” you need 5 years of game theory PhD. Hearing someone say “just trust me bro, the optimal strategy is that one” is not good enough. Voting was supposed to be for the masses…

drop everything to just start suing states and protesting for voting rights

I could get onboard with ranked-choice voting. My city used IRV for our latest mayoral primary election, and even though none of my ranked candidates won, I felt extremely satisfied that at least my voice was finally being heard. When a literal police-mayor got elected (winning primary by only 7000 votes), I had the comfort of full knowledge that this was not due to any spoiler effect on my part, but solely simply due to more people voting for him. If we’d campaign for ranked-choice voting in federal elections - presidential primaries and general - we can eliminate all the above hand-wringing. The Democratic party should be totally on board with this since they could finally get the Green protest vote.

So I am proposing that the Democratic party is acting irrationally and suboptimally, but you claim that the Democrats are acting most optimally, and it is the fringe left that is acting irrationally instead by refusing to accept a unfair split against all game theory guidance, causing all of us to eat shit (despite them making up only low single digits). Yet if the Democrats are so rational, how come they keep losing? Shouldn’t they have found an optimal strategy to get around the irrational ultimatum of the left? Yet here we are.