May 2024 – Spinor Info

Would you kindly…

Computer Games, Cybernetics No Responses »

May 272024

One of the catch phrases of the famous computer game, Bioshock, is “would you kindly”. It’s only near the end of the game that we learn that the protagonist is compelled to respond to this phrase and act accordingly. Presumably, omitting this phase would have had unpleasant consequences for the game’s antagonists.

I was reminded of this as I was playing with the “behind-the-scenes” setup instructions that I have for the language models GPT and Claude at my site wispl.com. The models are instructed on how to use tools, specifically Google (for searches) and Maxima (for computer algebra). I was perplexed as to why both models tended to overuse Google even when the conversation began with a question or request that should have required no searches at all.

The relevant part of the instructions sent to the chatbot at the beginning of a conversation used to read as follows:

“If your answer requires the most recent information or current events, respond solely with CSEARCH(query) with no additional text. For general queries or fact-checking that is not time-sensitive, respond solely with GSEARCH(query) and no additional text.“

In a moment of inspiration, however, I changed this to:

“If your answer requires the most recent information or current events, respond solely with CSEARCH(query) with no additional text. If your answer requires general queries or fact-checking that is not time-sensitive, respond solely with GSEARCH(query) and no additional text.“

Can you spot the tiny difference? All I did was to repeat the “If your answer requires” bit.

Problem (apparently) solved. The chatbot no longer appears to do Google queries when it doesn’t really need them. I just needed to make sure that the magic phrase explicitly accompanies each request. Much like “Would you kindly”, in the world of Bioshock.

Rising global insanity

History, Politics No Responses »

May 242024

I am watching HBO’s recent miniseries, The Regime, and I don’t know whether to laugh or cry. The satire is great. The allusion to historical or present-day political personalities, from Rasputin To Ceausescu, from Orban to Putin, are unmistakeable.

Yet it hits a little too close to home. OK, far too close to home. Which is to say, the world of 2024 resembles the fictitious world of The Regime a little too much. Authoritarians on the rise, wars of conquest, pardon me, “peaceful reunification”, meddling by China, efforts to undermine trust in the fundamental institutions of Western democracy… And frankly, sometimes it’s hard to tell the difference between the “real” CNN vs. its fictitious depiction in The Regime.

Indeed, I was just reading about Argentina’s new head of state, Javier Milei. El Loco, he calls himself, and it’s fitting: if Time’s article is to be believed (and it is certainly consistent with what I’ve been reading elsewhere), what we have here is a delusional conspiracy theorist with a messianic complex. Yikes!

But just in case nutty politicians in a distant land in the southern hemisphere don’t bother us, here’s another piece of news and this really gave me the creeps the other day. Dutch firm ASML, the world’s premier maker of extreme ultraviolet lithography equipment (i.e., machines that can make the highest of the high tech chips) assured us that it has the capability to remotely disable their machines, used by chipmaker TSMC in Taiwan, in case of a Chinese invasion.

I find the fact that ASML even thought it necessary to reassure us that this option exists deeply unsettling. It makes me believe that the question of the next Great War is not if, but when… and the most likely answer is, sooner than we’d like.

Seriously, can’t we just hand the world over to AI-assisted cats to manage? I am seriously inclined to believe that we’d all (cats and humans as well as machines) be better off that way.

Heterogeneous China

Culture, History, Politics, Religion No Responses »

May 202024

Occasionally I chat with people about China, in comparison with Europe. Many view China as a political entity that managed to maintain its unity even as Europe split into a great many competing states, often engaged in warfare.

I think this picture is patently false.

Sure, Europe has sovereign states, separated by borders. And sure, Europe has a multitude of languages, different ethnicities. On the other hand, while the Roman Empire may have officially ceased to exist many centuries ago, its unifying influence did not vanish. On the contrary, Europe’s traditions, customs, institutions, legal systems and religion, languages and writing systems remain dominated by the continent’s Roman heritage. Come to think of it, my own name (Viktor) is a Latin word!

And it’s not like folks in China all speak the same language. Look at this amazing map that I just came across, showing the multitude of languages spoken in different regions of China:

So what about those devastating wars, then? Europe was, after all, at the center of two world wars just in the past century, and there were plenty of conflicts (e.g., Napoleon) in prior centuries as well.

Yet all these conflicts are dwarfed by some of the wars the Chinese waged among themselves in the past. Wars with millions, even tens of millions killed were not uncommon, at a time when there were far fewer people in the world to begin with than in the 20th century. Measured by the death toll relative to the size of the human population, the two world wars of the 20th century were minor skirmishes in comparison with conflicts like the Three Kingdoms War or the An Lushan Rebellion.

So perhaps there are more similarities between China and Europe than at first meets the eye. We must not let the concept of state borders (themselves a relatively modern invention, a product of the Peace of Westphalia of 1648 that established the modern concept of state sovereignty) mislead us.

Microlensing black hole shadows

Astronomy, Personal, Physics No Responses »

May 062024

A couple of months ago, I came across a nice paper, by Verma and Silk (of Silk damping fame, as he’s known to cosmologists), showing what would happen if we had a chance to view the “shadow” of a supermassive black hole as it is microlensed by an intervening smaller black hole along the line-of-sight.

It occurred to me that I have the means to model this. At first I thought I’d write a short paper. But there really is nothing new that I can add to what Verma and Silk said in their paper, other than a nice animation produced by my ray tracing code.

So here it is. A brief animation of a small black hole passing in front of the famous “shadow”.

Things are not exactly to scale, of course, but for what it’s worth, this video corresponds roughly to a 10,000 solar mass black hole passing through, halfway between us and Sagittarius A*.

Oppenheimer

Film, History, Physics 2 Responses »

May 052024

I finally saw last year’s blockbuster, Oppenheimer. Let’s just say that my reaction to the film is not exactly in the mainstream.

That is, Best Picture my ass.

I am okay with Murphy’s Best Actor. Downey Jr. was especially good, earning his Best Supporting Actor in a role that I can only describe as unpleasant, playing the main villain of the Oppenheimer story, Lewis Strauss.

An actual photo of the real Oppenheimer

But the film?

For starters, there’s the jumbled timeline. I am deeply familiar with the Manhattan project, and reasonably familiar with Oppenheimer’s life, including the story of the humiliating revocation of his security clearance in the 1950s. Even so, I was confused: I had a hard time keeping track of what I was seeing.

Then, there are some of the portrayals. Teller was unrecognizable. Where was the famous limp? And what’s with the accent? Sometimes, no accent at all, sometimes an accent that, whatever it was, didn’t sound even remotely like Teller’s. For some of the other, well-known physicists, it was same thing: I’m glad the closed caption sometimes showed the name of the person talking, otherwise, I swear I would not have known that one of them was Szilard, for instance. And Groves? His portrayal by Matt Damon was more like a caricature than the real general.

And then there are the gratuitous sex scenes. I hope I don’t come across as a prude by mentioning this, but… was it really necessary? I mean, yes, I get it, their penetrating questions about Oppenheimer’s private life were metaphorically undressing him, but was it really necessary to assume that the audience is so dumb, they won’t “get it” unless you put Oppenheimer, stark naked, fucking his girlfriend right there in the chair in the conference room while he is being interrogated? Seriously, this was so over the top, I could not believe my eyes. My reaction was that they were trying to out-Kubrick Kubrick, but without the talent of Kubrick (and I am decidedly not a Kubrick fan.)

Then how about the conversations? Some of them, I swear, sounded like a bad AI (no, not GPT-4 or Claude 3, more like GPT-2 or compact versions of Llama) trying to recreate conversations between scientists. I don’t want to set an impossible standard here. How about just meeting the standard, say… of a sitcom? The Big Bang Theory and Young Sheldon are both more respectful of the science (and the intellectual quality of discussions between scientists) than this film.

And some of the scenes were just grossly inauthentic. Never mind misrepresenting the then-perceived significance of the Oppenheimer-Snyder paper on gravitational collapse (yes, it is significant, but no, the term “black hole” was not even coined until a quarter century later), what was that with that childish celebration when the print edition arrived? By then, Oppenheimer and his colleagues would have known for months that the paper was accepted. Oppenheimer would have seen, and corrected, the galley proofs. The fact that print copies of the journal would appear on the appointed date would have been neither a surprise nor news to anyone involved.

What about the things that were omitted from the film? And no, I am not talking about technical details, not even the massive role facilities other than Los Alamos played in the development of the bomb. How about Oppenheimer’s 1960 visit to Hiroshima? It could have offered some profound moments, perhaps even allowing the film to conclude in a way much more fitting than the stupid “burn the atmosphere” CGI.

And speaking of CGI… what’s with the Trinity explosion itself? I read somewhere that it was not CGI. I could tell… it felt cheap. A bit like the explosion of the planet Alderaan in the original Star Wars movie, before the remaster.

The film had some redeeming segments, especially in the final half hour, but even those were overplayed, like that final (as far as I know, wholly fictitious) conversation between Oppenheimer and Einstein. Certainly not enough to salvage the movie for me. The best part were the end credits, as the music score was decent (not sure about Best Original Score quality, but it was enjoyable).

All in all, between the two acclaimed blockbusters from last year, in my view, Barbie won hands down.

Incidentally, I reminded myself that I had an equally negative view of another famous blockbuster from ten years ago, Interstellar: crappy story, crappy science, a psychedelic scene that wanted to be a bit Kubrick-like but couldn’t quite make it (and I absolutely hated what Kubrick has done with the closing scenes of 2001: A Space Odyssey). What I didn’t realize until this moment is that both Interstellar and Oppenheimer were directed by the same Christopher Nolan. Guess that makes it official: I am no fan of Christopher Nolan! On the other hand, I suppose I am a fan of his younger brother: I liked Westworld, and I am beyond impressed by what he did with Fallout.