Honestly, we should start doing hardware-based age verification instead. Have the government run a simple yes/no service for individuals to be able to verify their age. The service simply asks if you’re over 18, and the government responds with a simple yes/no.

You verify your identity on the device once when setting it up, it asks the government if you’re over 18, and then your user account is verified as an adult when the “yes” response is returned. The only time it would need to be repeated is when someone turns 18, which would be something the user would need to manually prompt their device to retry. And notably, the government isn’t being pinged for every site you visit, they only got pinged for the initial device setup. So they don’t get access to any of your browsing data.

Now your phone can automatically send a “yes, I’m over 18” signal to any site or service that asks. And kids won’t be verified, meaning they won’t even be able to see the “are you over 18” prompts; they’ll simply be booted off the site (or in Discord’s case, restricted) as soon as it automatically asks their device for an age verification. No action is required on the user’s part, and the site/service didn’t need any invasive info about who you are. As far as an adult is concerned, they got direct access to the site without any kind of annoying “are you over 18” prompt. And as far as a child is concerned, they got automatically redirected right back to Google’s home page as soon as they clicked the porn link.

For shared devices (like computers) it could be handled on a per user basis. You verify your age on Windows/Linux/MacOS when creating the account, and then whenever you’re logged in, any site can simply ask if you’re over 18. Don’t want your kid to stumble across porn? Don’t verify their account. Now safeguarding kids on the internet is as simple as parents safeguarding their computer password and refusing to verify their child accounts.

It’s basically the best of all worlds:

• The government/private data brokers don’t get free access to your browsing data, like what would happen if every individual site asked the government for verification. This is our current reality, with data brokers hoovering up photos of IDs to feed to their data scientists.
• The adult user only needs to take action once to verify their age, and then after that the age gates are automatically opened. You don’t need to verify independently with each site, because your device handles that automatically during the initial handshake.
• Sites don’t get any additional personal info about you, except for the automatic pass/fail hardware response saying that you’re over 18. They don’t need to collect your info to pass to a third party verification system. They don’t need to ask the government, because that has already been done. And they don’t need to worry about things like GDPR compliance for collected info, because there is no additional collected info.
• Your browsing info isn’t shared with third parties, because the sites/services you use have no need to ask third parties for verification.

Of course it’ll never happen though, because it would restrict what kinds of info data brokers could collect and sell.

Yeah, self-hosting it can be a bear, especially since you need to deal with the whole “bots trying to kill it will regularly post CSAM in random channels, and if any of your users are in that channel it will federate to your own server and now you have CSAM saved on your server’s cache” stuff. It’s the same problem that Lemmy was dealing with during Reddit’s APIcolypse. You can always choose not to federate, but that largely defeats the point of the protocol existing in the first place.

You also need to set up TURN servers to get functional voice/video calls. WebRTC (like voice/video calling) tends to throw a fit without some sort of TURN functionality. That’s something the average Joe won’t know how to do, and is typically going to require a paid tier from some external host like Cloudflare.

Edit: I looked it up. Cloudflare offers TURN servers, with the first 1000GB for free each month, but then it charges for use after that. But that does mean a server that gets used for video calls more than a few hours per month could end up incurring costs. Because that TURN server would be handling all of the video streaming data, so it will quickly eat that 1000GB limit. It also means true self-hosting is prohibitively difficult, as you’d be tying yourself to an external provider unless you go out of your way to host your own TURN server.

I’m currently researching Matrix Synapse, to see if it would be worth the hassle to self-host a server.

And this is why generic EULAs should be heavily regulated, and allowed to be negotiated like any other contract. Allow me to pencil in a “you’ll allow me to uniquely watermark the scan of my ID so it can be traced back to this specific request, and agree to pay me $500M if that scan is ever included in a data breach or sold to additional third-party vendors” clause.

Oh, Discord doesn’t want to agree to that? Gee, if the company is deleting everything immediately and there’s no risk of a leak/intentional sale, what’s the harm in including it? How’s the saying go? Something about “if you have nothing to hide”?

It heavily depends on your instance, fwiw. .world tends to lean closer to Reddit in terms of moderation. And then .ml leans the other direction, and heavily moderates in favor of communism.

Yeah, the old “strict parents make sneaky kids” thing is unfortunately very true. You won’t stop the kids from doing it. You’ll just lose the ability to monitor them when they inevitably do it behind your back anyways.

Audiobooks mostly, and I know it’s not the same

FWIW, studies have found that reading and speech activate different language regions of the brain… But oddly enough, audiobooks activate the reading parts of the brain, not just the speech parts. So it may be more similar than you’d initially think.

I finally got my AudioBookShelf instance up and running last year, around March. In those remaining 7 months, I “read” (listened to) over 50 books. I used to be a voracious reader as a child. But at some point I lost the spark, as finding time to read got harder and harder. Audiobooks have reignited that love for reading, in a way I can’t even put into words.

I find myself looking forward to my daily commute, because it means I can listen to another two or three chapters in the car. I’ve even started taking the longer route home (which is technically more fuel efficient, but adds like 10 minutes to my commute) because I don’t mind the extra time in the car. I find myself wanting to wash the dishes or fold the laundry, because I can have my earbuds in while I work.

I still doomscroll on YouTube shorts or Instagram reels occasionally, but that doesn’t mean I have to give up long form content like reading. They fill two entirely separate niches.