Yes, rolling r sounds completely different to z-with-dot/rz. But if you start saying it and then move your tongue just a bit back on palate, rz comes out (or sz, if you stop the tongue vibration). So it is "anatomically" close, not close by sound.

Story behind rz is a bit different. In Polish we have letter "z-with-dot" which sound similar to french 'j' (like in "je" or "jalousie"). Way r is pronounced is Polish is close to "z-with-dot" tongue-wise (r is front of palate, z-with-dot is middle of palate, vibrating in both cases, even if slightly differently).
In old Polish, a lot of words had "r" in some places, which, due to language evolution (and probably people getting lazy, as rolling r is harder to say than "z-with-dot") slowly converted into z-with-dot. But in meantime, they were something in between r and "z-with-dot" and "rz" was used to signify that. These days, it should be just "z-with-dot", but it is a spelling mistake, because you know, poets from 150 years ago and old people.
For normal Pole, "rz" and "z-with-dot" is pronounced exactly the same. Same for "u" and "o-with-accent", and "h" and "ch". Some people will claim there is a small difference, but they are mostly snobs - more than 95% of people won't know the difference.
Useful ones are "cz", "dz", "dz-with-dot", "dz-with-accent", "sz", they really indicate separate sounds.

https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fen.wikiversity.org%2Fwik...

Let's start with Bitcoin and worry about AI later...

Can you provide the specific example of such text and questions? (if I'll find my own and get different results, you will claim it was too easy) I don't believe in _most_ of the questions wrong. Some yes, but over 50%, unlikely, for 1 paragraph text, for latest version of ChatGPT.

ChatGPT is not preprompted to pretend to be a human. It would have to be specifically prepared for that particular task, there are too many protections right now. As for "ability to understand and consider scientific ideas outside regurgitating 'the mainstream' narrative" - what percentage of humans do you think qualifies for that? AI doesn't have to be smart to pass Turing test, it just have to sound slightly smarter than average citizen. I think we passed that threshold already.

So many arguments anti-AI are based on 'it is not really creative', 'it just repeats combination of things it has seen before', 'there is no soul'... So do 90% of people in the world. Unless you want to deny their humanity (which I could agree with, if you want to go this way), don't diss AI on that basis.

If you go to their website (solreader.com), in the middle of the page there is a picture of glasses and if you click on Navigate, you get a closeup. You weren't much off - font seems to be 8px tall, but it is variable width, so a is around 5 pixels wide, which allows slightly more words in one line.
It is still pretty horrible quality-wise. Plus, their claim that they cost only 10% of apple headset, instead of saying 'we cost the same as Oculus Quest 2, at 10% of capabilities'...

Half life expectation was due to high children mortality. If you survived till 5, average lifespan was shorter than today, but not as drastically. Don't remember numbers offhand, but something like 60 vs 75, rather than 35 vs 75 if you count the children.

Can you provide sources on that ancient Chinese ban on using gunpowder as weapon? From Wikipedia on Gunpowder, as soon as it became widely known, all types of bombs and fire arrows were used. I think that guns were not created by China back then due to lack of technology or creativity, not any type of enforced ban.

But topic in question is metascam. If you compare crypto to selling you a car which is supposed to run infinitely on just air, no gas (only to later learn it doesn't work at all), these guys are robbing the transport full of those cars, stealing them and reselling on black market for a double price.

I agree with NFT being stupid, but 'fungible' means something else. It is not about mutability, it is about equivalence. From Wikipedia:

In economics, fungibility is the property of a good or a commodity whose individual units are essentially interchangeable and each of whose parts is indistinguishable from another part.[1][2] Fungible tokens are the ones that can be exchanged or replaced; for example, a $100 note can easily be exchanged for twenty $5 bills. In contrast, non-fungible tokens cannot be exchanged in the same manner.

So, from what I understand (and I don't understand too much about the actual quantum mechanism here), with perfect setup and measurements, they should win 100% times in case of 'no-reality' and 88% in case of 'reality'. They win 94% of the time, which means they are experiment is flawed by some kind of measurement/setup/whatever errors. 94% is 6% from 88% and 6% from 100%. Why do they think that this proves 'no-reality' case, rather than just saying 'we have no clue'?
Is it the case that possible errors can only decrease chance of winning, never increase it? It would then make some sense.

I have written this in confusing way. I meant rms rejecting Javascript, while Taler is explicitly catering for it. Replace his/he by rms in last sentence.

He pushes for using GNU Taler, but from the Taler doc:

For easier browser-side processing, we restrict some integers to the range that is safely representable in JavaScript. Subset of numbers: Integers in the inclusive range 0 .. (2^53 - 1).

So, even if his baby gets adopter by the world (big 'if'), he will be in same position of not being to pay online, if he rejects javascript...

From reading the abstract, it seems it is a catheter, so it has 'infinite' length. Only metric here is the width, which is 2mm OR 1/10th of the inch for the metric-impaired.

It is not a self-propelling robot going thought your lungs. It is just a very thin catheter, which is too thin to allow direct articulation, so they are driving it with the help of external magnetic fields.

In theory it is very compact. Most digital storage technologies are either 2d, or just multiple layers of 2d. With DNA you get a string of linear data, but it can pack itself in dense 3d space.
I think that practical difficulties (access time, error rate, lack of durability, etc) will make it mostly useless anyway.