What are you reading, watching and listening to? - All languages

[Raphael]

Interesting, Travis, but what does that have to do with my content warnings about a book I'm reading in a non-academic context?
.
.
.
.
.
.
.
Anyway, back to the book I'm reading. Here's a quote that particularly caught my eye:

Tuesday, November 5 [1929], was election day, and the market was closed all day. In the New York mayoralty race, the Democratic incumbent, James J. Walker, scored a landslide victory over his Republican opponent, F. H. La Guardia, who had been soundly denounced by the Democrats as a socialist.

Sometimes, the past really is a foreign country.

[fusijui]

Very belatedly --

Navn Ukjent was a great character, although I think they generally copped out on the Mesolithic people's language - you'd think that after several years in Norway, at least some of them would move beyond speaking Pidgin Norwegian*), and the Vikings speak Norwegian just fine when they're not speaking Old Norse among themselves (if it were only about Navn, that'd be fine; he is clearly cultivating an image).
*) I assume that's what they speak in the original; I watched the dubbed German version, where they speak Pidgin German.

I think some characters spoke one of the Saami languages, too, at least briefly -- that may be somewhat anachronistic or very anachronistic, depending on what 'original time' they came from. But that's beside the point; I also think the portrayal of the 'Mesolithics' language-learning problems was overplayed. (For the record, I watched the series with the original soundtrack but with English subtitles... and I'm getting pretty deaf, and I had much the same impression as you.)

The second series was a disappointment to me, when it introduced the Conspiracy Theory model of storytelling that, it sometimes seems to me, is the norm in modern(-ish) settings lately. I would have been very happy to watch more blocks of episodic, character/setting-based dramas about 'a pair of fantastically miss-matched cop buddies', but lost interest pretty quickly when it turned into some reality-hopping fantasy show.

[hwhatting]

I think some characters spoke one of the Saami languages, too, at least briefly -- that may be somewhat anachronistic or very anachronistic, depending on what 'original time' they came from.

As far as I remember, the Saami is spoken in settings having to do with the Norse characters, so the main question would be if they used a reconstruction of the relevant Saami Variety of ca. 1000 years back or just a modern Saami variety (I guess the latter, but I don't know Saam and wouldn't be able to tell).

The second series was a disappointment to me, when it introduced the Conspiracy Theory model of storytelling that, it sometimes seems to me, is the norm in modern(-ish) settings lately. I would have been very happy to watch more blocks of episodic, character/setting-based dramas about 'a pair of fantastically miss-matched cop buddies', but lost interest pretty quickly when it turned into some reality-hopping fantasy show.

To be fair, elements of that were already there in the first season, but they really cranked it up in season 2. I agree that less of that would have been better.

[fusijui]

The second series was a disappointment to me, when it introduced the Conspiracy Theory model of storytelling that, it sometimes seems to me, is the norm in modern(-ish) settings lately. I would have been very happy to watch more blocks of episodic, character/setting-based dramas about 'a pair of fantastically miss-matched cop buddies', but lost interest pretty quickly when it turned into some reality-hopping fantasy show.

To be fair, elements of that were already there in the first season, but they really cranked it up in season 2. I agree that less of that would have been better.

I think you're right; I probably just didn't recognize the signs in the first series while I was watching it. If there's a third one, I'll probably watch it too, regardless of my disappointment!

Also, I meant to say earlier that I really, really liked the title sequences -- the commute-to-work through beforeigner-populated Oslo, under Bobby Bland's theme song. I don't know why it worked for me, but it sure did.

Anyway, next on my agenda: re-watching Da Vinci's Inquest and Utopia (the British version); meant to do the latter earlier, when COVID was a hotter topic, but didn't get around to it.

[hwhatting]

Also, I meant to say earlier that I really, really liked the title sequences -- the commute-to-work through beforeigner-populated Oslo, under Bobby Bland's theme song. I don't know why it worked for me, but it sure did.

Yes, me too. I especially liked that they filmed a new sequence for each episode, tying it into the plot for each episode, instead of using the same sequence throughout the series or season. And the song, which I hadn't know before, really fits; I had it on repeat on my Spotify for a couple of days after watching the series.

[Raphael]

One last note on The Great Crash: At one point, a 1920s cabinet member is quoted as referring to Minneapolis as being in the "Northwest". Interesting that perceptions of geography were so different so recently.

[Linguoboy]

One last note on The Great Crash: At one point, a 1920s cabinet member is quoted as referring to Minneapolis as being in the "Northwest". Interesting that perceptions of geography were so different so recently.

I live in the territory of the "Old Northwest", which didn't even include Minneapolis.

The usage in that book seems like a relic of a time when "the West" began at the Mississippi River.

[fusijui]

Starts at the Mississippi? My mother, born in Philly, thought she'd be seeing the Rockies and buffalo when they moved to Ohio as a kid -- it was 'way out west' for their family/community.

[Moose-tache]

That's just run-of-the-mill regionalism. cf. this famous New Yorker cover, showing the veiw of the world looking west from 9th Avenue in Manhattan. The inhabitants of major northeastern cities have the additional excuse that they've been told their whole lives that their community is the center of the universe.

The real test is whether people from that place identify it the same way. It was once the case that Minnesotans would say they live in the West or the Northwest. There is a university in Illinois called "Northwestern," which is now east of the geographic center of population in the United States. Ohio was called part of the "Old Northwest," but after the Civil War I think this kind of talk was only done in retrospect, like "Ohio was settled when George Washington traveled to the Old Northwest," or some shit. I doubt anybody in the 20s was still calling it that.

Back on topic, my wife is obsessed with Sky Castle, the latest Netflix K-drama, which is a whole genre now. As someone who lived for the better part of a decade in Korea, I have never been able to get into K-dramas. Does anyone here see the appeal?

[fusijui]

Back on topic, my wife is obsessed with Sky Castle, the latest Netflix K-drama, which is a whole genre now. As someone who lived for the better part of a decade in Korea, I have never been able to get into K-dramas. Does anyone here see the appeal?

I do see the appeal. But (a big but) only if the drama in question is about something I'm already invested in -- e.g., The Grand Chef/Gourmet/Sikgaek (I like cooking), Leverage (I was very fond of the American original... though I'm not sure the remake really falls in the 'K-drama' genre). I've tried out others based on good reviews or possible 'investment' (historical dramas, for example) but -- never finished them. So, I guess, I'd say -- if it's a drama that you'd be interested in, Korea can deliver the goods; if it's not, it's just as dire and tedious as one from any other country. Being into them because they're K-dramas? I really can't see it.

[rotting bones]

Postcapitalism and How to Stop Fascism by Paul Mason made me hate humans slightly less. Thank you, Marxism!

Read some other books like Nine Kinds of Naked, Uncommon Sense and Waiting for the Galactic Bus.

---

Interesting YouTube and other free online videos:

NCommander: https://youtube.com/playlist?list=PLCpe ... fkO8vLF9t5

javidx9: https://www.youtube.com/playlist?list=P ... ZZ_4IgteYf

Evidence that reason works, disproving what basically seems like all popular political ideologies of the 21st century: https://youtu.be/feYv2wZFbb0

NFTs: https://youtu.be/YQ_xWvX1n9g

Rewatched Louis Theroux's Israel documentary: https://vimeo.com/102569427 (At least I think this was the one.)

Secret Space Force: https://youtu.be/EYvnKc908Fw

Gravel Institute

I couldn't find Odin's Song on YouTube for a while. Here it is again: https://youtu.be/-mhLAt6xubY

---

I tried some new paid streaming services:

Hoichoi, the Bengali streaming service. I had been told to watch Byomkesh Season 5. The most famous Bengali detective is Feluda, but Byomkesh deals with more adult themes. The writing is also dense with Sanskrit vocabulary. I had forgotten how much they praise the intellectual caliber of fictional detectives. It's bothering me for some reason. I also watched the unrelated thriller Tiktiki.

BritBox has tons of Agatha Christie murder mysteries including Poirot.

---

Playing Calcudoku, Sagrada and Hive. Trying to get back into playing Shogi at the amateur Dan level. Still stuck at 5 kyu.

PS. Forgot to mention I'm intermittently reading the Hugo Award winning story Barrayar.

[rotting bones]

a) We're not very good at what computers do best, ie. run algorithms very reliably and very quickly.

These things depend on architecture. Mechanical decision-making is called an algorithm regardless of the architecture: https://drive.google.com/file/d/11FiA9l ... p=drivesdk

b) Computers aren't very good at a lot of what our brain does. (With a humongous load of data, a few months of work and constant nursing by a few engineers, you can teach a computer to recognize cats. Babies pick up that skill pretty much instantly.)

The most powerful artificial neural networks we have are about as complex as a frog's brain IIRC.

Singulitarians conclude that a) means our brains are really a very sucky type of computer and that b) will be solved by throwing more transistors at the problem.

I think a more realistic conclusion is that the brain isn't really very much like a computer at all.

My conjectures are based on the current state of science, which accepts the validity of the Church-Turing thesis. If you can prove that models of problem-solving exist beyond the Turing machine, you could publish it and immortalize your name. Until then, it's not rational to bet on it.

The view of such a simulation is oddly dualistic. The brain simulation is somehow completely separable from the human body; you can jury-rig a connection to a VR environment or a camera or something. I tend to think you'd have to emulate a complete human body as well, and sensory inputs to match.

People survive all kinds of damage to the body that would kill them if it happened to the head. Besides, if the rest of the body is less complex than the brain, it wouldn't be as expensive to simulate it.

From a fictional premise, I tend to think it doesn't make for terribly interesting stories. Once you get back the technobabble and the dreadful earnestness of it all, you end up with the same old robot story.

I like that pony story because it carefully highlights how singularitarianism is motivated by social alienation.

[Ares Land]

My conjectures are based on the current state of science, which accepts the validity of the Church-Turing thesis. If you can prove that models of problem-solving exist beyond the Turing machine, you could publish it and immortalize your name. Until then, it's not rational to bet on it.

The Church-Turing thesis is about numerical problem solving. Is the human brain's function really numerical problem solving?

[Travis B.]

My conjectures are based on the current state of science, which accepts the validity of the Church-Turing thesis. If you can prove that models of problem-solving exist beyond the Turing machine, you could publish it and immortalize your name. Until then, it's not rational to bet on it.

The Church-Turing thesis is about numerical problem solving. Is the human brain's function really numerical problem solving?

The question is whether there is anything about the human brain that could not be simulated at a physical level by a sufficiently powerful computer (emphasis on sufficiently powerful). If one cannot theoretically build a computer of some sort (whether conventional or quantum) that can do so, then it would imply that there exist physical systems that simply cannot be simulated. If you prove that, you probably will win a Turing Award. Of course, it may be practically impossible given the physical constraints of computing to do so in any reasonable amount of memory or time (e.g. it may belong to a complexity class in memory or time that renders it highly intractable), but that is another story. Again, if you prove that, you might win a Turing Award.

[Ares Land]

I think there are several open questions here. First: can all physical processes be simulated by a Turing machine? If there are such processes, are they involved in the human mind?
I don't think the question has been settled either way so far. (On the plus side, we all get awards either way!)

Assuming that the human mind can be simulated by a Turing machine (sounds reasonable) and physically possible (on the face of it, that sounds reasonable too), it may turn out that simulating a human mind to any interesting degree of precision is impractical given present-day or near-future technology.

It may also turn out that the human brain is about as efficient as it gets in terms of cognition and any attempt at emulation will turn out suboptimal.

I think the immediate limit (without even approaching the Church Turing hypothesis) is our understanding of the brain. As far as I know, we have only the vaguest idea of how neurons actually work, or what neurotransmitters do (besides a general idea of what happens if there's a shortage of dopamin or if you serotonin into the system).

[Travis B.]

It may also turn out that the human brain is about as efficient as it gets in terms of cognition and any attempt at emulation will turn out suboptimal.

It may very well turn out that the only practical way, given space or time constraints, to emulate a human brain is to essentially build an artificial human brain at the nanoscale, i.e. a specialized computer designed specifically to simulate the human brain that directly approximates the form and function of the human brain. That said, I do not personally believe that there is any theoretical reasons why one can't do that - if it can be done in wetware at the present, it can be done in a physical approximation of said wetware.

I think the immediate limit (without even approaching the Church Turing hypothesis) is our understanding of the brain. As far as I know, we have only the vaguest idea of how neurons actually work, or what neurotransmitters do (besides a general idea of what happens if there's a shortage of dopamin or if you serotonin into the system).

I think this is the real problem presented by trying to emulate the human brain - we simply do not know enough about how it functions to do so, and do we have the means to acquire the knowledge necessary to do so (as all the means that we have to peer into the inner workings of the human brain are crude and highly inexact)?

[Zju]

Late to the train and even later to the discussion, but has anybody else read The three bodies problem / Dark forest? I so wish the books lacked the implicitly compulsory ccp propaganda sections author's impressions of various nations' inherent manners; the lack of editorial review really shows. Both books would the about 40% to 50% shorter and that would be a major improvement.

[zompist]

My conjectures are based on the current state of science, which accepts the validity of the Church-Turing thesis. If you can prove that models of problem-solving exist beyond the Turing machine, you could publish it and immortalize your name. Until then, it's not rational to bet on it.

The Church-Turing thesis is about numerical problem solving. Is the human brain's function really numerical problem solving?

The question is whether there is anything about the human brain that could not be simulated at a physical level by a sufficiently powerful computer (emphasis on sufficiently powerful). If one cannot theoretically build a computer of some sort (whether conventional or quantum) that can do so, then it would imply that there exist physical systems that simply cannot be simulated. If you prove that, you probably will win a Turing Award. Of course, it may be practically impossible given the physical constraints of computing to do so in any reasonable amount of memory or time (e.g. it may belong to a complexity class in memory or time that renders it highly intractable), but that is another story. Again, if you prove that, you might win a Turing Award.

Shades of comp.ai.philosophy! I spent so many hours there 30 years ago. Looks like the arguments of today are identical!

Well, except that today, I think, people assume that an AI will be a Deep Learning type neural network that no one can understand, rather than an arcane LISP algorithm that only the programmer understands, but forgets to comment.

I think Travis is right here, but kind of trivially, meaning no awards for anyone. Could you simulate a human brain using a computer the size of the universe? I don't see why not! Can you do it with one the size of, oh, a human brain? Not guaranteed at all. All these hifalutin references to Turing just come down to that: you can do it with an infinite-sized computer. That hardly addresses Ares Land's point about efficiency. (Y'all do remember that Turing machines are infinite? And that Moore's Law will never get to infinity?)

I don't think anything magic or supernatural is going on inside the brain, but I do think analog and sensorimotor processes are way more important than AI types usually acknowledge. The brain is a machine, yes, but a machine for operating a physical body in the physical world. More like a robot, then. But people think of robots as having programs entirely separate from their machinery, and that just doesn't seem to be the case for the brain. If all you want to do is identify faces for the cops, that probably doesn't matter, but maybe it's a bad way of thinking about the problem if you really want an omnifunctional android.

(A rather minor demonstration of this is that animal brains often solve geometric problems with brain geometry. That is, there's an array of neurons that precisely corresponds to a geometric array in the world. This allows some easy processing of certain problems. E.g., if neuron A represents point A', and point B' is next to that, where is the corresponding neuron B? Right next to A. This makes things like edge detection easier. It also allows insects to easily perceive polarization patterns with their tiny brains. Yes, you can do all this with math. But using physical geometry is a pretty nice hack.)

A big unresolved question is qualia. Before joining your consciousness uploader, I want to know: does it preserve qualia? If the researcher says "Oh yes, we even stored it in a bigint!" then I'm not jumping in. This doesn't mean I think qualia is an unsolvable problem, just that AI has no theory at all of qualia and therefore no solution.

[zompist]

Late to the train and even later to the discussion, but has anybody else read The three bodies problem / Dark forest? I so wish the books lacked the implicitly compulsory ccp propaganda sections author's impressions of various nations' inherent manners; the lack of editorial review really shows. Both books would the about 40% to 50% shorter and that would be a major improvement.

I reviewed the first book here. I haven't read past that.

I don't recall anything like CPC propaganda. There's a Chinese point of view, but it would be rather weird for a Chinese guy to write sf that didn't have that point of view. I'd also note that Liu is extremely critical of the Cultural Revolution, which is in line with Deng but probably borderline heretical for Xi.

[Travis B.]

I think Travis is right here, but kind of trivially, meaning no awards for anyone. Could you simulate a human brain using a computer the size of the universe? I don't see why not! Can you do it with one the size of, oh, a human brain? Not guaranteed at all. All these hifalutin references to Turing just come down to that: you can do it with an infinite-sized computer. That hardly addresses Ares Land's point about efficiency. (Y'all do remember that Turing machines are infinite? And that Moore's Law will never get to infinity?)

BTW, determining that a computer, conventional or quantum, with unlimited resources can simulate a human brain isn't something I'd expect would win one an award. As you say, it is the kind of thing that is almost trivially true. What I was saying is that the opposite, determining that a conventional or quantum computer with unlimited resources can't simulate a human brain, would likely win one a Turing Award.