You were really getting shafted as a web developer and likely could’ve made a lot more… But on the other hand I’m glad you found something you like more. I still prefer programming, with or without AI assistance, to just about any other work I could do. I have a gig where I generally get to solve slightly different problems every day (very wide domain) and I have no idea what other job could provide this without being in danger from LLMs.

I know a few carpenters of the self-employed variety and I gather it can be a very rewarding job if you’re into it. I could never, I need more stimulation for my ADHD, I need to be able to do a week worth of work in 2 hours when things are overdue, it doesn’t really work in carpentry lol

They’re getting cheaper actually. The gigantic touchscreen is cheaper than a bespoke interior design with physical buttons for every model.

Of course that’s not reflected in the price

The end result is what other people see and what gets you paid. That’s why people want to do it via AI.

Personally, I could sink 10k hours into whatever visual art and I’ll never be good at it. I don’t enjoy it either. I can only hire an expensive artist or use AI if I want something made. Since using AI art sucks, I’ve opted not to continue work on the video game I started work on. An alternative is to pay for asset packs from a studio that sells them, but then it’s just an asset flip.

The batteries aren’t cheap. That’s still the expensive part of an EV, even though it’s nowhere near as expensive as it used to be.

But overall, the same issue plagues all new cars. There have been very few great innovations in the last decade or so, so now they just have to cram more techy tech into cars to get people to buy a new one. I mean no real innovations aside from the electrification itself of course.

Go look at high-end cars in 2016. Carplay/Android Auto was already a thing, though optional in most cars, and not available in all yet. Safety features like automatic braking, adaptive cruise, lane keep assist were already available. Massage seats were old news. There was just nothing new to add to a car to differentiate it from others. Tesla had just come along a few years ago with a slick UI and over the air updates, and people were blown away with it. What do you do? You start imitating the success of Tesla, going too far (too much tech) instead of integrating only the good things (actually responsive screen, for example).

Hell, I maintain that you can get 90% of what you’d ever want from a car, if you’re willing to own ICE, from a well-equipped 2005-2013 Mercedes S-Class by upgrading it with a AA/Carplay module. There’s little in its 2014-2020 successor that makes it truly a better car - only a slightly nicer infotainment system, better transmission (even that only in the facelift) and different (some better, some worse) engine options. So in 2020, with the new model, they went crazy with screens, making the interior actually uglier than the old ones. The electric equivalent, EQS, was even more screen-focused, only because it came out a year or 2 later, not because it’s electric.

Once we get to cheaper cars, however, it’s even harder to add any real new innovations. MB’s Magic body control (camera controlled hydraulic suspension that knows when to soften up for a pothole) is super expensive. 4 wheel steering is expensive. A large touchscreen is not expensive. A sim card for online services is not expensive.

Yes, turns out the Adam Sandler movies get him and his friends a nice vacation to e.g Hawaii or some South African resort AND a paycheck to go along with it. When you realize that, you kinda understand why he’s producing that slop rather than genuinely good films. The slop is consistent, it’ll basically always do well enough to justify the cost.

Mi Scusi guy is also the Straight To Jail guy from parks and recreation, which I would NOT have realized without looking at his IMDB page.

AI has no search algorithms. It’s about 1000% less efficient, because its algorithms are built for token prediction, not sifting through websites.

It doesn’t need to; It just relies on the normal web search to get better data than what it could generate from its training data alone. This wasn’t a thing with the early popular LLM interfaces, which is why they were ridiculously bad. They still make plenty of mistakes, but at least now you get a link to what it looked up (so you can facepalm when you see it’s way too generic and your question was very specific, so the generated answer is still wrong)

If you hardly realize they’re even there, you’re more likely to click on them since most advertisers pay to be shown on relevant searches, not random ones.

E.g if you run a pet food store, you’ll let Google show your store whenever pet related things are searched for. And that’s where having a blog comes in handy, you can have it show articles relevant to the search rather than just the store front page.

Many people have noticed a decline going back to 2020 when Google decided they need to show more ads, which means you need to search more times or check more pages than the first one rather than getting the desired result right away.

Kagi seems to be better at this since it’s paid and doesn’t need to show you ads.

But of course you’re right that content itself has sloppified too.

Oh yeah, Firefox used to take years to change major version number, then they changed to… what was it now, a 6 week schedule? To keep up with Chrome’s versions.

It’s not as one-sided as that comment makes it seem (see my reply to them). Many corporations do contribute back or fund development, especially since it’s easier and cheaper to do that than to maintain a hard fork for all their changes. Hell, with Linux, despite GPL, they could still run a bunch of proprietary code on your device through modules since those don’t have to be compiled into the kernel and thus aren’t bound by the kernel’s GPL.

With something like a video game console, if your whole operating system is easy to replicate, that makes it much easier for someone to build a copycat OS that lets anyone run that console’s games on non-sanctioned hardware. Wouldn’t need an emulator if you just straight up get the whole OS running on wholly different hardware and modify it to your liking easily and you could probably skirt copy protection too. Hell, it would be GREAT for us consumers. Not so great, however, for companies that spend tens or hundreds of millions developing a new console hoping it will sell tens of millions of units to pay off (and then get a cut on the games themselves).

Instead, we’ve arrived at a sort of hybrid model, where companies contribute to the development of things that aren’t specific to their hardware or whatever they believe their advantage to be, while contributing to things that help everyone, but won’t reveal their “secret sauce”.

Release engineering team lead at FreeBSD commented on reddit that Sony would actually fund a bunch of development work that went into FreeBSD proper rather than just their fork, but do it super under the radar (mostly because they don’t want competitors knowing what they’re up to in future versions, particularly in terms of hardware support). Netflix contributes tons to BSD and uses it not because of the license but because of the networking stack, as they wouldn’t have to disclose shit if they did run Linux, it’s all on their servers. Apple funded development of LLVM (which many other large corporations have also contributed to).

Source on the Sony and Netflix part, can’t link directly to the comment because new reddit ui sucks and I can’t access old without logging in anymore

I could be wrong, but I don’t think that’s what the question was about.

It was “if people are trying to use Anthropic or OpenAI’s services to build a profitable company, why wouldn’t Anthropic or OpenAI just build that company themselves”

And the answer there would be lack of subject matter experts. You need people who know the industry they’re building the product for if you want to have any fighting chance. AI can work as a force multiplier, trying to use it as a completely autonomous agent that gets near-zero human feedback is hopeless. I doubt any of the AI companies want to take on thousands of employees just to use their own AI to take risks on new industries when they could just sell the AI to people who take on the risk.

Hey, any day can be a Taco Tuesday if you want it to be. Especially if the taco in question is himself the kind that changes things names on a whim. You could have 7 or more Tuesdays per week for all I care.

Not like you have to specify “I’m making this reservation to celebrate Trump’s death” though

Orban

He was replaced with Peter Hungarian. No, I’m not kidding, that’s what the new guy’s last name (Magyar) means in Hungarian.

Can’t wait for Trump’s reign to be over and John American to take over.

As you’ve been both fit and technically obese, I suspect you also realize exercise is much harder to do when you’re obese. Not just because moving your body is harder, but because you have less energy and thus it might be much harder to motivate yourself. It’s a vicious cycle where gaining weight is even easier when sedentary and being obese makes it harder to not be sedentary.

And with socialized healthcare in most countries, everyone’s paying the price. I say this as someone currently technically obese as well.

Luckily most people don’t WANT to be obese, so GLP-1 inhibitors are a bit of a godsend. Forcing anyone to lose weight I can’t condone, but giving people the best possible chances to be healthy seems like a no-brainer.

People were also severely stressed in the centuries past (remember, the 8 hour workday is a recent invention). Less stress helps for sure, but the availability of calorie-dense food engineered to be delicious is probably even worse.

Did you use the phone at all? 2 days of idling, to me, implies never turning on the screen or opening any apps

Removed by mod