This unearths some thoughts, back from my English degree twenty years ago
But basically I don't think those problems are all meant to be solved
Yes, the supernatural events described on stage are very difficult to pull off even with modern effects. But they were impossible with Jacobean setups. And that's kind of the point
There is always a tension between what is seen and spoken in that play. We're meant to get the sense that Macbeth is partly talking himself into his frenzy
Supernatural events happen but they kind of have a bathetic, eye-rolling quality to them, like the prophecy that "Great Birnam Wood to high Dunsinane Hill / Shall come against him"
Which YEAH, it does, in that Macbeth's enemies construct a bunch of battering rams from the forest timber. You are supposed to groan at that. Magic is way weaker than any of the human beings and their ordinary vices
> the supernatural events described on stage are very difficult to pull off even with modern effects. But they were impossible with Jacobean setups
Speaking generally, as a former theatre professional, using onstage "effects" is less effective than firing the audience's imagination, and inviting them to do the work - which, of course, Jacobean artists were well able to do, and their audiences arguably more prepared to accept. The best plays - and I do think Maccers is among them, although the three productions of it I've been in were mediocre to dire - give huge scope for that.
This was a wonderful article, that nails what's wonderful - and frustrating - about the show.
>Which YEAH, it does, in that Macbeth's enemies construct a bunch of battering rams from the forest timber
My memory might be failing but I thought it was that the soldiers were disguising their movements using branches cut from the trees so it gave the impression that the trees themselves were moving towards Macbeth's castle.
I have found this extremely relevant as a (primarily) non-verbal thinker.
I don't, generally, think in words, more in - I guess I would call it something like meta-shapes? A sense of a shape but not things I can exactly visualise.
(You might be surprised to read this and then hear I have an English degree. Surely I thought about Shakespeare in words?! Nope. Shapes, movement, structures)
For me, having to write is critical because it is the only way I practice serialising my thoughts in a way other people can understand.
If I do not then I get very "deep" into my own way of sensing ideas and it's difficult to dig myself back out.
This might also be why I have never been very enchanted by LLMs? They only seem to "think" verbally. So it is always a translation effort for me.
I never can really enter any "flow" state with an LLM. My intuition is that highly verbal thinkers can enter flow with LLMs very easily
Interesting. I have designed code that way. And the shapes aren't UML diagram elements or anything like that, they're just... shapes. I'll slowly walk around, in the hall or outside, and be kind of seeing these shapes and vaguely moving my hands around as I sort out the relationships between them.
I think I have produced reasonably good designs. Don't ask me to teach anyone how I do it, though.
I find this and the parent comment highly relatable with the caveat that I also find it extremely intuitive and rewarding to get good outputs from quality LLM's like Sol or Astra, and I haven't had any trouble with "flow state".
One of the most rewarding things for me is figuring out a good shape for a system and how it would interoperate with the other systems, especially in a way that reframes other parts of the codebase in a way that bring clarity and makes it more intuitive to work with. Creating the right ontologies can make all the difference in what you can do with a project. It's a form of creating mathematical objects.
For example, a Unity game I work on has quest and dialog systems driven by visual scripting graphs. We had two way dialog with different units for player response choices and npc dialog. But we wanted to expand to letting NPC's have dialog with each other as well as conversations with more than two participants. I went outside and thought it over, which largely amounted to visualizing a dialog node graph and a feeling in the back of my mind like it was trying to perform a kind of geometric shape-fitting exercise. A fitment solution jumped out at me to have only one "Dialog" node shared by all participants, with a "participant" value on it. If the player parses this node then the options go on-screen as responses, while if an NPC parses this node with multiple options in it, it picks one. And this lets you voice the player if you want, and enables some things like overhearing other NPC's talk to an NPC then talking to that NPC yourself and having the same tree.
And for quests, the quests had just been for the player, but I was thinking about how to make scripted events in-game easiest to work with for script team who primarily works in visual scripting. Similar story - let the NPC's have their own little quests, with task stages, which are easy to track and make branching choices from, and let the NPC's definition for how to use that quest contain a collection of actions to override the typical actions available to it, so an NPC in a specific "quest" can't do things you don't want it to do, a common enough case that it's preferable to making a series of conditions on the general action planner like "not in quest A"
And timing myself, it took 1-2 hours each time to write out the detailed plan for how I wanted each thing implemented in the game with some other tasks thrown in, and it paid off after Astra worked on it until it was done. It was awesome coming back to something pretty much exactly what I asked for each time.
"I don't, generally, think in words" is not the same as "I don't think in words".
Like I already said, if thought was exclusively in words for humans, humans wouldn't have the "It's on the tip of my tongue" problem. It's blindingly obvious that thought does not occur exclusively with words.
+1... this is me as well. Using LLMs in a conversational mode does not fit my thinking. I use AI help in my editor through targeted code generation, explanations, etc. and I'm writing my own harness to hopefully get a better feel for the shape of LLMs that way. How are you adapting?
This is one of the things that did in Jan Hendrik Schön.
I can't remember the exact measurement but he was showing a set of curves that were simply too neat for a relatively small sample size - roughly 100 or so plastic transistors. In reality you cannot get that neat an effect without a huge sample population
> What I can't wrap my head around is, businesses are run by people, and all these entrepreneurs are making explicit decisions to abuse their fellow human beings
This is a thought I want to pull on
I have a view that organisations, with their network of misaligned incentives, mean that companies act something like an alien lifeform. Something with a decentralised brain, unable to plan beyond immediate stimuli
Companies consume and self-cannibalise. They are both aggressively rational in pursuit of profit but irrational when it comes to preserving their ecosystems
Many companies look like groups of people where nobody has total control. Everyone is reacting to some stakeholder, even the CEO is terrified of "the shareholders". So these organisations cut off healthy limbs (layoffs), engage in lurid rituals (AI), and wastefully destroy the natural resources of the humans who work for them
I first heard this idea in a James Bridle lecture, where he suggested that the strange behaviour of companies was a good starting point for understanding what an alien lifeform might behave like. And he observed the ways that humans and machines now mirror each other.
Inside our society, controlled by these aliens, humans act like machines (Taylorism at work, self-censorship for algorithms at play). Machines act like humans (AI, spam, computer programs mediating relationships).
Underneath it is a frantic need to quantize everything so it can be dominated
I remember recently they polled workers about AI pressure: "the AI doesn't work very well, but my boss wants to see me using it." They polled the boss: "I didn't say they had to use it, they're choosing to use it and this'll look good to the executive". The executive: "well this costs a bunch but it'll look good to shareholders and we should encourage more use so it'll look even better." The shareholders: "do we look like we give a shit?"
Out of all of them, I'd assume the shareholders to be wrong.
Look at the state of the world. Yes, you do look like you give a shit.
Also there's an effect here where if just 10% of shareholders care about this it still makes sense to go for it if you think you can be faster than your competition.
Corporations are too complex for any single person to truly control. Attention of leadership necessarily dilutes with each additional layer of management. The result is this weird corporate chaos where nobody is completely in charge because everybody has to take charge of something, however small, to progress, or the whole could accomplish nothing at all.
The result is this flailing tentacled monster where each appendix seems to have a mind of its own and sometimes littlw regard for even the other parts of its own body.
This reminded me of an old documentary about corporations: "If we look at the corporation as a legal person, it exhibits all the characteristics of a psychopath" https://www.youtube.com/watch?v=s5hEiANG4Uk
Yes you are right. With HTTP/2 responses can be sent out of order, but once they are sent down the pipe they have head of line blocking and have to arrive at that order, unlike QUIC.
Websockets don't really have an advantage here because typically websockets are implemented server-side as event based systems, and if a RPC paradigm is put on top of it has the same properties as HTTP/2 request/responses ordering. So if HTTP/2 doesn't have head of line blocking, then you can't argue that websockets do.
You are being downvoted but I think this is correct. I doubt Netflix is really doing a double blind RCT on which languages produce better AI pull requests. How would that even work, have two versions of each service in different languages?
It's anecdata and maybe, MAYBE, a spreadsheet. Or a Google Form somewhere.
I work at a large devsec company which uses primarily Go and TypeScript
I've found that the LLM generated Go has few mistakes, and generally isn't too obscure. But the volume of code is so high, colleagues do a bad job of reviewing it.
I've seen a lot of very silly decisions made, like returning the wrong HTTP code, or miscategorizing a metric used for an SLO, that I just don't think is helped by the sheer volume of code one has to wade through.
Ironically, we are considering migrating some initiatives to Rust, exactly because experiments indicate it works well with LLM development.
I'd argue that this is neither down to the programming language nor LLM used, but down to your own company's requirements, testing, and code review practices - but you did point the latter out so you're aware of it. I think the only thing you realistically can try to do is slow down.
But it's also a trap I fall into myself - I don't usually thoroughly review my own merge requests because I wrote them, but now that LLMs generate them I don't review my code thoroughly either, and some issues have fallen through. Nothing critical, just stuff that, if someone asked me to review it, I'd flag up as something that could be polished a bit.
Why would you think so? Rust provide better type system and abstraction mechanisms compared to Go, hence equivalent system should have less lines of code, at least in from my experience.
Yes, and that's the case I've seen as well. I would also say rust is also more heavy on symbols, which can make code look kinda hard to parse in places.
But weirdly Opus (N=1) in Claude Code does okay on it. Enough I can reliably have it write software and feel confident it works.
But basically I don't think those problems are all meant to be solved
Yes, the supernatural events described on stage are very difficult to pull off even with modern effects. But they were impossible with Jacobean setups. And that's kind of the point
There is always a tension between what is seen and spoken in that play. We're meant to get the sense that Macbeth is partly talking himself into his frenzy
Supernatural events happen but they kind of have a bathetic, eye-rolling quality to them, like the prophecy that "Great Birnam Wood to high Dunsinane Hill / Shall come against him"
Which YEAH, it does, in that Macbeth's enemies construct a bunch of battering rams from the forest timber. You are supposed to groan at that. Magic is way weaker than any of the human beings and their ordinary vices
reply