maxsilver

> Does anyone watch scheduled television anymore except local news and sportsball?

Yes, especially for folks not at home. It’s common to see multicast used in places where internet bandwidth is constrained, but network maybe not so much.

For example, maybe you have an older hotel, or a cruise ship or a hospital or a retirement home. It’s expensive to just buy better internet backhaul, and it’s expensive to retrofit those rooms, and you can lose a lot of money for every day the room is down.

But, with multicast, you can use the legacy internet connection and the legacy wiring in the room, but easily/cheaply ensure every room has a working TV with full broadcast TV or satellite TV, and anyone watching that is not pulling 5mbps to 10mbps from your internet backhaul.

It also all works from a single remote control, and it doesn’t ask you to log into anything, it doesn’t need your cell phone for anything, etc.

It’s niche, but there’s a number of places where it makes sense.

Yeah, it quickly becomes a "which came first, the chicken or the egg?" thing between junior devs and LLMs.

I will say, if you use a Mistral model, and if you insist your CSS framework is Bulma (tell it, 'no tailwind', 'no preprocessor'), it does okay at staying away from Tailwind. (Not perfect, not great, but okay).

No LLM I've used can handle raw CSS well (yet). If you are carefully curating your own classes and styles, you might just be on your own for a bit.

> these bots really want to just barf out tailwind-style crap.

I get it. The LLMs struggle most with state. They don’t have a real fix for that yet. People generally compensate by shoving everything into context, and making the context window as large as possible, which half-works.

Tailwind happens to be “stateless” CSS framework. Nothing uses anything else, nothing is shared, nothing is reused, nothing stacks. It’s super easy to write, since you don’t have to worry about anything else, and the styles are all duplicated dynamically and ‘compiled’ — to the point you can copy-and-paste a HTML block with tailwindcss classes from anywhere into your site, and it mostly ‘works’).

—-

Tailwind is uniquely suited for LLM use, because the problem Tailwind solves is the problem juniors (and now, LLMs) struggle with most. An LLM can happily write up a bunch of styles, without knowing any of the rest of the project state, and if it’s tailwind, it will mostly sort-of work.

It just also happens to be bad practice, this style of development is the exact thing we told everyone not to do for two decades. (“Inline styles are bad! Duplicate styles everywhere is bad! It’s bloated, it’s inefficient. It’s the mark of inexperienced front end. Don’t inline styles. Unless it’s a tailwindcss class, you can inline those styles, they get a pass I guess”).

We used to measure our JS and CSS in kilobytes, by 2011 standards this would be “far too bloated for production use”. For the old-timers, it can be hard to grapple with the idea that we’re just purposefully doing ‘worse’ front-end intentionally now. The calculation changes when half your content/styles/front-end is LLM-generated, and therefore completely disposable. Very “they don’t make them like they used to” vibes.

> Importantly, I think AI companies are motivated towards the overengineered solutions as they increase the buyer's token spend.

Yes that, and also, the more complicated the solution, the more likely no one reads or reviews it too carefully, and will instead depend on an LLM to ‘read’ and ‘review it’

Even ignoring token costs, there’s a high incentive for LLMs to generate complex solutions, because those solutions generate demand for further LLM use. (You don’t really want to review that 30,000 line pull request by hand, do you?)

> Also, when did we stop liking to learn?

When the economy got so bad for so many people, that every waking moment has to be either chasing fresh cash (or spent in recovery from cash-chasing, worrying about new cash), to the point they have to largely ignore their own long term goals or basic morals or principles.

You can blame all the new gadgets (phones/social media/tiktok/‘dopamine-things’) — but it’s a very much blaming the symptom, not the problem.

(It’s the meme. “Guys, this isn’t funny. Humans only do this when they’re very distressed”)

> iMessage is taking twenty minutes to sync a message between your laptop and your phone sitting six inches apart.

The iMessage one is super common, and is Apple's fault. Easiest way to reproduce it is to have two Macs. (got a desktop and a laptop and use them both? Chances are high you'll encounter it).

The HomeKit (via HomePod mini) is also super common. (HomePod Minis just have bad wifi and unreliable connections, there's something about their WiFi setup that's different from all other Apple products). It doesn't help that Apple spent years prioritizing HomePods as the HomeKit base (though they eventually fixed that, and let you assign an Apple TV to do it).

The others are also common, but not necessarily always Apples fault, as far as I can tell.

(the AirPods, for example, tend to go wherever 'most recent' sounds happen, but a lot of developers are unintentionally triggering conflicting behavior around this. Have Outlook open? An email notification will sound an alarm, stealing AirPod focus away from your other device, but the sound effect will already be done playing by the time your AirPods connect, so to the user, it just seems like the AirPods switched devices for "no reason".)

(HomeKit, for example, is supposed to support Eufy cameras. But Eufy cameras are garbage, despite having a large dedicated base station dock running 24/7, they can support only one small video stream to one single device, ever. So if you have two Eufy cameras installed, HomeKit will fail on the cameras constantly, but it's because of Eufy's basestation limitations, so it's not clear to me how Apple could 'fix' that)

---

The more Apple moves outside of it's own internal ecosystem, the more complex the interactions get, and the less control Apple can feasibly exert over the product lifecycle, so the more it starts "Microsoft-ing" it's work. (We joke about Microsoft Copilot, but Apple has five different products all named Apple TV, the Apple TV (hardware device), Apple TV (the TV software app, which runs on Apple TV, and iOS, but also on Roku and other SmartTVs), Apple TV (the storefront for buying movies and TV shows), and Apple TV (the subscription service) for watching Apple TV (the studio creating original content shows and movies, one show of which is actually called "The Studio")

> Honda was never in the large SUV game.

(The Honda Pilot and Honda Passport stare at you, with deep resentment)