I don't know if the comment was referring to this, but recently, people have been posting stuff about them requiring their new hire Jared Sumner, author of the Bun runtime, to first and foremost fix memory leaks that caused very high memory consumption for claude's CLI. The original source was them posting about the matter on X I think.
And at first glance, none of it was about complex runtime optimizations not present in Node, it was all "standard" closure-related JS/TS memory leak debugging (which can be a nightmare).
I don't have a link at hand because threads about it were mostly on Xitter. But I'm sure there are also more accessible retros about the posts on regular websites (HN threads, too).
After some experience, it feels to me (currently primarily a JS/TS developer) like most SPAs are ridden by memory leaks and insane memory usage. And, while it doesn't run in the browser, the same think seems to apply to Claude CLI.
Lexical closures used in long-living abstractions, especially when leveraging reactivity and similar ideas, seems to be a recipe for memory-devouring apps, regardless of browser rendering being involved or not.
The problems metastasize because most apps never run into scenarios where it matters, a page reload or exit always is close enough on the horizon to deprioritize memory usage issues.
But as soon as there are large allocations, such as the strings involved in LLM agent orchestration, or in non-trivial other scenarios, the "just ship it" approac requires careful revision.
Refactoring shit that used to "just work" with memory leaks is not always easy, no matter whose shit it is.
Code quality never really mattered to users of the software. You can have the most <whatever metric you care about> code and still have zero users or have high user frustration from users that you do have.
Code quality only matters in maintainability to developers. IMO it's a very subjective metric
I once used it to check whether it would identify some birds that are prevalent in my area.
Not related to the app's fubctionality, but it was pretty funny when I replayed my recording of parrot noises to crop it and the next moment, a walnut shell dropped from the tree above.
I can relate to writing being a substitute for rubber-ducking to a certain degree: often I write stuff down solely for this purpose.
Impersonating the reader used to be (still is?) an essential skill requirement for writing prose. And prose here is really a substitute for "informal technical writing".
But now you can vibe the same app 1000 times for root beer, coca cola, ginger ale, even a milkshake, and nobody will ever have to have a new idea again!
I also miss the all-capitals ARM spelling. I think they've never been the same since they've changed that, since around the same time their business strategy went from sensible to nonsense.
Not in British English, is my point. Look for example on the BBC, it's written "Nasa". I'm not saying I agree with it, but it's something you learn living in UK and former commonwealth countries.
The BBC (all caps) style guide splits the treatment of acronyms; BBC is all caps as it is spoken as a series of letters Bee Bee Cee.
Nasa is spoken as a single word.
From the BBC:
Use the abbreviated form of a title without explanation only if there is no chance of any misunderstanding (eg UN, Nato, IRA, BBC). Otherwise, spell it out in full at first reference, or introduce a label (eg the public sector union Unite).
Where you would normally pronounce the abbreviation as a string of letters - an initialism - use all capitals with no full stops or spaces (eg FA, UNHCR). However, our style is to use lower case with an initial cap for acronyms where you would normally pronounce the set of letters as a word (eg Maga, Sars, Aids, Nasa, Opec, Apec).
The ABC Australia style guide is a hot mess (first examples: Quantas TAFE modem are inconsistent "spoken words" and other issues) ... I'm wondering if it was last edited by an LLM or young intern.
Edit: looked into it and the first paragraph doesn't exhibit any LLM "tells" to me, so I'd rather read it in full or research about the source than judge it. Leaving the rest of my comment because it is my opinion on the argument of using LLMs to rewrite text.
I don't know if this was done here.
=====
I haven't read TFA, and this explanation comes up again and again, but I'd rather read broken English (or German), than the "enhanced" version.
Considering that LLM rewriting using non-specialized tools is more often than not far from preserving intent and meaning of any input, I'd say I think this applies even more for non-native speakers.
You wouldn't say "maybe the author is not a physician, so they might have used an LLM to fill in the Latin terms and medication doses" or "not a scientist, used ChatGPT to do the statistics using my notebook of empirical data" either.
Language has value and simple language or slightly wrong grammar is preferable to a verbose and glossy distortion of the input.
Sorry if this doesn't apply, since I didn't click the link.
And yeah I'm sure my comment is verbose and partially wrong in my English, but well.
Totally agree, my point was that I didn't get the impression that the article was LLM-generated, for that reason. The commenter I was replying to seemed to think the article was obviously LLM-generated, so LLM-aided translation was one possible explanation, but I don't have any particular reason to believe that's what the author actually did.
I've read the first paragraph rather than skimming it now, and it does show LLM tells, and not so few as to appear accidental...
:D
> water everywhere was not only a necessity but also a marker of status, a matter of discipline, and often an aesthetic pursuit. That’s why, when you look closely at the story of coffee in the Ottoman world, you don’t find only roasted seeds, copper cezves, and foaming cups—you also encounter an unexpectedly refined culture of water. Even today, as specialty coffee digs into water hardness, alkalinity, and pH, it’s tempting to think that some of our “scientific instincts” are, in a way, echoes of the same land.
water everywhere was not only a necessity but also a marker of status, a matter of discipline, and often an aesthetic pursuit. That’s why, when you look closely at the story of coffee in the Ottoman world, you don’t find only roasted seeds, copper cezves, and foaming cups—you also encounter an unexpectedly refined culture of water. Even today, as specialty coffee digs into water hardness, alkalinity, and pH, it’s tempting to think that some of our “scientific instincts” are, in a way, echoes of the same land.
but yeah I still didn't read it all or the think about the source, the website is unknown to me though.
> I just don't understand what's even intended by this.
I might be misinterpreting, but according to the landing page, this is the intention:
> Personal Computer gives Perplexity Computer and the Comet Assistant always-on, local access to your machine's files, apps, and sessions through a continuously running compact desktop.
> It's a persistent digital proxy of you. Controllable from any device, anywhere.
That being said, the grandeur and bombastic language also seems fitting for something less sinister, like an even worse version of MS Recall maybe? Combined with, let's say... agents!
That's it! You Personal Computer is your agent and not only may act on your behalf, it also communicates your preferences and intentions.
Yes of course. Unfortunately many of those decisions get distorted and captured by bad actors, creating a reasonable skepticism.
If you care about solving climate change: instead of yelling at climate change denialist you should direct more effort into advocating for policy and messaging that acknowledges and mitigates the harms while keeping you expect people to endure
It doesn't sound to me like you contribute any valuable argument that would improve the "PR" for the goal of protecting environmental living conditions for humans though.
I'm mostly too old to go to clubs, but to me techno and other electronic music is not about stages or visual effects, quite the opposite. Well OK, VFX can be cool.
But getting lost in music, in a darkened room with some intentionally disorienting VFX; or simply none, loud electronic music in a room with many people is already quite an experience...
that's quite different from being at a festival or at a show like this, which looks more like a musical opera performance to me.
For big room EDM, was there ever a time when it was not about laser shows etc?
I mean there's nothing wrong with stage shows, pop music and lasers.
And at first glance, none of it was about complex runtime optimizations not present in Node, it was all "standard" closure-related JS/TS memory leak debugging (which can be a nightmare).
I don't have a link at hand because threads about it were mostly on Xitter. But I'm sure there are also more accessible retros about the posts on regular websites (HN threads, too).
reply