JohnEdwa

Apple spending 3 years researching and determining it is not doable due to technical limitations:

It’s only a matter of time until they completely ban porn.

LLMs are the current big buzzword and the main ones that “don’t work”, because people assume and expect them to be intelligent and actually know and understand things, which they simply do not. Their purpose is to generate text in a way that a human would and for that they actually work perfectly - get a competent LLM and a human and ask them to write about something, and you are very unlikely to spot which one is the machine unless you can catch them lying, and even then it might just be a clueless human talking about things he kinda understands but isn’t an expert of. Like me.
But they are constantly being used for all kinds of purposes that they really don’t yet fit well, because you can’t actually trust anything they say.

Image generation mainly has issues with hands and fingers so they aren’t bullet proof at making fake realistic imagery, but for many subjects and style they can create images that are pretty much impossible to identify as being generated. Civit.ai is full of examples. Most people think it doesn’t work yet because they mostly see someone throwing simple prompts into midjourney and taking the first thing it generates for an article thumbnail.

And image identification definitely works, but it’s… Quirky. I said it can’t be used to identify mushrooms, because nothing can identify two things that look exactly the same from one another. But give one enough photos of every single hot wheels car that exists, and you can get one that will perfectly recognize which one you have. But it will also tell you that a shoe or a tree is one of them, because it only knows about hot wheels cars.
Making one that is trying to identify absolutely everything from a photo, like Google Lens, will still misidentify some things as the dataset is so enormous, but so would a human. Just that for an AI, “I don’t know” is never an option, it always says the most likely answer it thinks is right.

You are lumping a whole lot of different things that work in completely different ways under the singular label of AI, and while I can’t really blame you as that is what the industry does as well, image recognition, image generation and large language models like chat-gpt all work entirely differently.
Image recognition especially can be trained to be extremely accurate with a properly restricted scope and a good dataset, but even so it would never be enough for identifying mushrooms because no matter if it’s being done by the perfect AI or an organic meatbag, mushrooms simply cannot be accurately identified from a single picture as they can look literally identical to one another in many ways.

And parrots totally can learn what words mean. Just like how a dog can learn what “Sit”, “Paw” or “Let’s go for a walk” mean, parrots just also have the ability to “talk”.

And even back when Microsoft bought the mobile phone operations the company making the phones was Microsoft Mobile and Microsoft only leased the brand name from Nokia to use on their mobile phones for 10 years - same as with HMD Global today.

Patents last for 20 years, that’s a long time for something unique and groundbreaking to become mundane and seemingly obvious in hindsight, especially when almost everything these days builds on top of something already existing at a break neck pace.

But the problem with the current system is that everybody has to try to patent absolutely everything they come up with because if they don’t somebody else might and then sue you for it, and instead of the patent offices actually doing their jobs and dismissing them outright so they would be free to use for everyone, they grant patents on the most simplest or broadest of things.

The silver lining is that plenty of great new things have been made specifically because people have been trying to avoid someone else’s patents - “necessity is the mother of all inventions”, literally.

Also when you combine this with some other news, like “Bots now make up nearly half of all internet traffic, and that’s very bad news for our security”, it skews it even further to being rather meaningless - bots are probably doing quite a bit less bittorrenting.

There was a blip in time a few years back when you really didn’t need to learn how to bittorrent, you could just google what you wanted and you’d get a pirate streaming site showing it. I’ve been torrenting for almost two decades now and still kissanime/aniwatch/zoto whatever was the faster choice most of the time, and even now I sometimes use the Kodi plugin Otaku when I just want to watch an ep or two while eating, and that grabs stuff from streaming sites. Granted it is getting worse and worse, with the sites being absolutely riddled with ads and seemingly rebranding twice a month due to takedowns which has me going back to nyaa more often.

Ah, downloading postage stamp sized anime releases that still took all day. Forget binging a series in one go, you watched an episode or two a day because that was as fast as you could get them.

But you can’t forget the absolute minefield of the era of Kazaa, Emule & Limewire - you never knew when you’d get a virus, something random, literally just cp or actually manage to grab the thing you intended to.

Also check if you have TMC drivers and if they have Stealthchop enabled, it can cause these types of issues. Especially for the extruder motor.

The question wasn’t “Is it efficient or cheap”, it was how much energy is in a battery, and if and for how long would it run a fridge. If you also want to add one more point to why you probably shouldn’t do it, car starter batteries don’t generally like to be deeply discharged, you’d want to get a marine battery for that use.
As for how much the inverter would cost, depends on the fridge, but Amazon has a 1000W inverter for around $85, that should be enough for most. Ours could run from a 300W one, they cost around $30. Pretty handy devices if you want to run any kinds of electronics from a car anyway, I have one for when I want to charge my laptop and RC batteries on the field.

Extrusion variance and bad tolerance filament will also exhibit the exact same looking issues, not just poor Z alignment.

Just to be sure, you have z-hop disabled? That can cause all kinds of issues if the gantry isn’t absolutely perfect.
Another thing would be making sure all variable line width printing options are disabled, I.e ones that try to fill gaps with thicker lines or print thinner lines faster to stretch the filament.
And finally, melt temperature oscillations, making sure every feature - outer/inner walls and infill - are printed at the exact same, slower, speed.

As for actual z issues, if the bed/gantry isn’t freely moving and has even a slight bind on the screws/axis, it can miss microsteps. One rather harsh troubleshooting tip is to reduce z-microstepping to a lower value, as that gives it more torque per step and usually you do not need the z to move anywhere near as gradually as some microstepping setups theoretically might allow.

Watt hours are watt hours. Sure the compressor won’t run on 12 volts as is but the energy is there, just needs a converter.

Fwiw, our 15 year old fridge uses around 1000Wh per day.

Humans had to learn that the hard way in multiple places, unfortunately.

There are still plenty of stubborn people that cling to Windows 7, Steam dropped support a few months back when they upgraded the… Electron version, I believe? Had something to do with chrome/chromium removing win 7 support.

The logic behind the concept originally made sense, they manufacture just one car with all the features as that reduces manufacturing overhead by a ton, much more than what they would save by having one with heated seats and one without (especially when multiplied by all the possible configurations), but instead of only providing the model at the price point with all of them enabled, they disable some for the cheaper models - this is possible because car prices aren’t really based on how much they actually cost to manufacture.

This then lead into allowing people to pay to enable the features later if they wanted to, because why not, they are already there. Iirc Tesla was one of the first to do this with unlocking range, performance and “self-driving” stuff.

And finally it morphed into a subscription option because hey, if you only need heated seats a few months a year, why pay for the others? Only $10/month! And $15 for that, and $5 for that, and…

Same goes for this Audi, the subscription is an option if you buy the lower spec model and then later don’t want to pay the full price to enable the features permanently.

It’s been fun abusing a Lemmy bug (or an API feature, depending on how you look at it) that lets you see what a removed comment originally said on a certain .ml instance. Unsurprisingly, they almost always were something negative about China or Russia.

The moderation effort required to clean up these ads must be massive.

And we’ve mostly hit the limit of usable maximum sizes. For like the last two decades you could upgrade your TV to the next bigger size every few years for the same money you paid for the last one.
I remember starting with a maybe… 21" LCD TV back in 2005ish, and for that money today I could get like 70" TV. I don’t have space to fit one that large, nor do I have any need for it even if I could.

We haven’t exactly reached this state because of some big, singular fluke nobody could have predicted, there hasn’t been some asteroid impact or super volcano eruption that suddenly messed the planet up.
Humanity as has known about the issues of pollution and climate change for over a hundred years, known how dire the situation has been for decades, and has still done mostly nothing productive to end up here.

It would have taken a few generations of people all making massive changes willingly to fix this, but almost none of us did. I think we just can’t care enough about the future when there is now and here to experience, and it’s just finally the “find out” phase after the “fuck around”.