241–250 of 261 posts
Earlier quoted context omitted.
They won't actually show you who said what though. Twitter trialed that feature then they quickly turned it off after everyone realized half of the maga influencers were russians. It also kind of stinks because not every mistake should be immortalized and recorded forever. Blackmail and all that. It kind of ruins the internet in a different way.
influencers were russians ... or just hiding behind a VPN that exits there?
Earlier quoted context omitted.
That's defeatist, although certainly what they'd like. The world is not the US. Even if NSA captured and logged all internet traffic, they'd still only get a fraction of the information within without breaking all the encryption. And even if they could break the crypto, the ability would only have any power if/when they acted on it. Which in turns reveals the capability to both normal people and other nation states.…
Ok? So they use it when they feel like and not willy nilly? That doesnt change my perspective. They still have the ability to do it at will.
No, when they can get away with it. When the pros for them outweigh the cons. And every person affects that calculation in some small way - by their tolerance to injustice, by their outspokenness, by their ability to detect it, by the probability of an employee becoming a whistleblower and even by people's willingness to punish bad actors extra-legally.
> They still have the ability to do it at will.
Yes and your neighbor has the ability to wait on your doorstep until you open and then slit your throat.
He generally doesn't because of internal safeguards (humans have evolved emotions like remorse, though not all of them feel it - see psychopathy/ASPD for details) and external safeguards (the state generally opposes people infringing its monopoly on violence but in the absence of a state, he would still risk being hunted down and tortured/killed as punishment by your relatives and friends).
---
I get what you're saying and I oppose dragnet surveillance but seeing things in pure black and white does not help.
Earlier quoted context omitted.
I'm just surprised people use Chrome at all. Google has proven over and over they can't be trusted and will exploit you every chance they get.
95% of people who use Chrome have no clue what browser they are using. They got Chrome when it was bundled with every single installer ever for about a decade (which was so prolific and scummy that Microsoft had to make the "default app" picker system more defensive , because Chrome was abusing it more than microsoft apps were). When you installed Java, you also got Chrome set as your default browser with no interact…
Earlier quoted context omitted.
> Heck, most of them think the Internet is Chrome. In the end Google has achieved something that Microsoft couldn't with Internet Explorer, and won the Browser Wars. Google managed to aggressively advertise their browser by optional install "offer" within Windows installers of software. And they were aiming exactly at all those who couldn't tell the difference between the web browsers and who were conditioned by more…
What annoys me the most about this is that we as web developers have been here before with IE and managed to get out of it. Then we immediately forgot that lesson and helped Chrome be the new IE in a new coat by adopting ”modern web API’s” that were in fact proprietary Chrome API’s with low respect for the standards process, until every other browser was forced to accept it as a standard. All those ”Safari/Firefox ar…
It seems to me that adding AI to desktop apps and sending the data back to the mothership for processing is an amazing way to collect data from people who, for the most part, would be completely unaware it's even happening. Heck, most of them think the Internet is Chrome.
I'd argue that even pre-AI the average internet surfer never thinks about all the data the sites they use collect. I'm not even sure if my mom uses any apps that aren't web apps (maybe MS word). But for it all to go to one place? That's a scary amount of data.
It seems to me that adding AI to desktop apps and sending the data back to the mothership for processing is an amazing way to collect data from people who, for the most part, would be completely unaware it's even happening. Heck, most of them think the Internet is Chrome.
According to the current shortage of computing power and electricity, I suspect that what they really want is not your data, but the computing power and electricity from your device. If users' behaviors can be pre-labeled on their own devices, processed with AI, and then sent back, it might save a significant amount of internal computing costs.
https://chromeunboxed.com/chrome-memories-history-journeys
A 4B model is not very useful for text generation, but useful for classification, so you're probably right.
Earlier quoted context omitted.
I'd argue that even pre-AI the average internet surfer never thinks about all the data the sites they use collect. I'm not even sure if my mom uses any apps that aren't web apps (maybe MS word). But for it all to go to one place? That's a scary amount of data.
It already does, given how much of the internet is fronted by Cloudflare. They're better positioned than most to know what users are doing.
It seems to me that adding AI to desktop apps and sending the data back to the mothership for processing is an amazing way to collect data from people who, for the most part, would be completely unaware it's even happening. Heck, most of them think the Internet is Chrome.
I'm just not sure how this gives me control of my information, whether I want it sent or not to Google, and if they're retaining it for training or not. That last question I don't even want to ask because the first two doesn't seem clear. This could be simply fixed by adding the feature, and defaulting it off, and letting people learn about it and enable it.
That's not the goal. Turning your information into their information is the goal. If training an LLM on some data isn't copyright infringement but instead makes it a brand new non-derivative work, then training an LLM on your personal information arguably means it's no longer your personal information, but instead a legally distinct work.