Another option is to use software like VeraCrypt to encrypt sensitive data into a file on your PC. Make it look like an innocuous system file and put it in the system folder somewhere. Don't leave VeraCrypt installed - it will run installer-free, download it at your destination and remove it before traveling again.

There's a double level of security if you use the option where it has a second dummy decryption container in the same file that you can give away.

But yea, if you can securely download what you need, the ultimate choice is to not have the data with you at all.

How can Microsoft install something? Does it have free will, agentic rights, useraccount [sic] on Windows?

Microsoft provides a mechanism for monitor manufacturers to trigger installs. LG, a monitor maker, took advantage to install some adware.

TVs already show ads. They give you enough functionality that you want to hook them up to the internet.

HDMI ethernet already exists, and DDC/CI is an I2C bus. No reason you couldn't upload arbitrary data over those, but you'd need driver support on the computer side.

A monitor will work with default drivers. A mouse will work with default drivers. A keyboard will work with default drivers.

It's as if we have standard protocols for basic functionality of all the devices required to give permission to install drivers for non-essential devices.

Somebody didn't read the summary.

I don't think you know what model collapse is either. I don't mean that pejoratively, the whole concept is a muddled mess from the name on.

"Model collapse" isn't anything unique to AI. Any model will drift if it's repeated trained on its own output. Models are all imperfect fits and that imperfection compounds if they're not anchored to actual data. Put a different way, the data processing inequality says you can't create information through computation. It's not "collapse," it's drift.

Once upon a time I was registering images to each other. One of the techinques is to create an average across all of the images, register each individual to the average, and repeat. If you just do that naively though, the average (i.e. the model) will drift farther and farther from the mean of the original images, or even anything recognizable as being related to the original images.

"Model collapse" isn't a problem you fix. If you want to model human language you train on human language. If you want a self-evolving language model then you train on successive generations and see where you end up. Humans don't "get a form of model collapse" we experience exactly the same thing. We call it "languge evolving", accents, fads, "why are the kids so weird," etc.

People talk about model collapse as if AI generated training data is some kind of poison because it's AI generated. It's not. If you train on human vetted data, whether or not it's AI generated, you'll get something that humans are okay with. If you train on SEO optimization web pages that have never been read or written by any human, particularly if you do it generationally, then you're going to get drift, whether or not the stuff is AI generated, random, or whatever.

So they're taking all your data, selling it, and making you pay for the privilege? Must be a nice business if you have zero scruples.

Why do you need OnStar? Who would want to steal a GM car? All the GM brands are at the very bottom of Consumer Reports reliability index for 2026.

OnStar uses cellular service to connect, so if you have an emergency, either both OnStar and your cell phone work, or neither work. So there's no advantage to using OnStar rather than just dialing 911. 911 services are allowed access to your phone's GPS location regardless of settings.

It just seems like a pork package to sell along with the clear coat. Not to mention the theft of your data without compensation, whether you subscribe or not. You have to physically disable the modem to avoid that.

Tesla is awful due to the lack of physical controls where they would really make sense. It is an actively driver-hostile car. Other than the powertrain, which is amazing, the car is completely obnoxious. The door handles in particular are a huge problem and have been banned in China. Why would you intentionally make door handles that are super awkward to open rather than just pull? Why would you make someone dig through a menu to adjust the mirrors, or adjust the A/C vents and other controls? Trying to shift with the right stalk? Awkward and horrible. I think they thought it was sexy to just have a screen and nothing else, but in reality it just absolutely sucks. Before you say, "but have you owned one?" Yes, yes I have. It's my wife's current daily driver. And I avoid driving it like the plague.

Yeah, totally unprecedented....

https://www.youtube.com/watch?...

Hey, our product is classified as a weapon by the US military....

Maybe not, but they do want to use the super advanced powerful a-awesome best AI that can do things like that when it's "without production classifiers" that make it safe.

You still don't understand what model collapse is.

Sure, and we here on Slashdot are criticising the silly article. And I'm criticising your silly repeating of the silly claim that polarization is not a quantum phenomenon because you can "explain" it (in a very narrow energy range if you squint and don't look too close) using classical physics. I probably wouldn't have replied at all if you hadn't told one of your other replies "it's okay if you don't understand quantum mechanics."

The actual paper doesn't use "first quantum material" at all. It says "first class of room-temperature quantum materials that are intrinsically sensitive to the quantum statistical properties defining all forms of light". They actually write it all out twice, which suggests to me that they were aware that shortening it to "quantum material" would sound ridiculous.

In fact, on casual reading it appears to be a fancy polarizing filter that's also sensitive to coherence.

It's pretty straightforward. Instead of simple counting on your fingers you treat your right thumb as a 5, your left fingers as 10s and your left thumb as a 50, which lets you count to ten on one hand or a hundred on both. It's fairly obvious once you understand digital notation, but you could also teach the concept using it, just like an abacus.

If you really wanted to get carried away you could treat your fingers as binary digits and count to 1024.

The systems aren't creative they're just really really fancy search engines that regurgitate what they have ingested.

Are we still saying this? These systems have solved a wide variety of open math problems some of which were borderline famous.

But what are the practical implications of that? Isn't it just as likely that those math problems haven't been solved because there was not any particularly good reason to solve them?

'Dem goal posts are a walking.

I figured if I read the comments on this post, I'd find all the incels and rednecks whining about Fark. The truth is, it hasn't really changed much, other than the fact that the people running the site grew up, and the site grew up with them and shed that frat-boy mentality.

They also had to get rid of adult content due to Google becoming the morality police of the internet when it comes to sites that they will allow ads to be run against. They are in turn pressured by various right wing lobby groups in the United States, who also pressure credit card payment systems not to support adult content. Blame the religious right for that. A legacy of our puritan origins. (Fark tried moving all the Boobies links to foobies.com, and that experiment ultimately did not work.) They have to police comment photos as well for that reason.

Fark is still a terrific link aggregator with funny headlines, and no AI or algorithms to filter it. The traffic has fallen off a good amount from its peak years, but it's still very lively.

I also find it funny that people here are complaining that fark is 'woke' or whatever, when I see plenty of borderline socialist ideology on display in posts every day on Slashdot. If you are complaining about that, why are you here? Yea, Fark is certainly anti-fascist, and I'm great with that.

I'd say Slashdot is in a much more degraded state than Fark. The original sense of community on Slashdot is long gone, though it still provides a useful aggregation of privacy and security focused tech news, which I appreciate. The comment sections here are far nastier on average. Fark is still fairly civil other than a few trolls here and there, and it's easy to block them.