So you think you have solved your age verification problem with zero-knowledge proofs... https://www.eff.org/deeplinks/2026/08/zkps-arent-age-verification-silver-bullets
@evacide "ZKP" aaah you mean Single Source of All Knowledge
or Single Point of Failure
or High Value Target
or Firewall of China as a Service
or...
@evacide I'm a bit sad that the article doesn't really explain any details and just skirt over "ZKPs bad", there's no explanation why ZKPs proof calculation would decrease privacy, there's no explanation about why the CA authorities would learn about a proof being emitted. No real demonstration about the centralization effect of using ZKPs.
@evacide Age verification is a total bs.
What if...
1. Stiff legal consequences for large media sites that fail to quickly respond to reports of bullying or predatory behavior. (thinking of Roblox)
2. Funding to help schools and public libraries create ad-free well moderated spaces FOR young people.
I know the bans and age checks aren't really about solving the problem but what WOULD be?
It's just a lot of "go away" and "not for you" ... where will kids go exactly?
Whenever I talk about making social media for kids and teens I really struggle to get anyone excited about the idea. Which I think mirrors the way that there are no physical "third places" for young people in a lot of the country.
Then everyone wonders why teens hang around in parking lots and lie about their age and go on horrible websites.
If we build it they will come.
@futurebird @evacide You'll need to reinvent/resurrect concepts of "moderation" that have largely disappeared from the zeitgeist, imo.
You'll need a much MUCH more active moderation team that is actually present socially, and serving a basically-full-time-role being visibly present in and stepping-in-where-necessary to social situations, as well as specifically soliciting and uplifting behavior (especially creative/effortful activities) from kids that you want to promote a culture of.
Prior art you can look at Neopets -- not just the forums themselves, but the Neopian Times, e.g., and all the other adult-engaged participatory activities.
I think CricketMag (Muse Magazine, etc.) also had spaces like this.
@futurebird @evacide But I think there's few enough examples that no internet site should be the primary model. Instead I'd suggest looking at sleepaway camps, I think?
In terms of camper-counselor ratios, in terms of additional staff with specific behavioral specialities that can be reached to as a resource/esclation by the counselors, in terms of explicit whole-team trainings before beginning to be in charge of the kids. (... possibly in terms of specifically assigning kids info groups with specific "their counselors"? I don't know what's possible/gained/lost there.)
... except, without a "beginning of camp"/"end of camp" chronological frame and even more significantly without an off-season. (At least facilities costs would be lower?)
@evacide i remember those german corrupt hackers in the times around 2015 manipultaing iPhones and switching the hashes for the fingerprint for people trying to unlock. They controlled put off devices as well all the time. Camera, microphone, processes running, just the display was off and all that could indicate a being put on. Complete fooling!
When hardware is insecure, nothing is secure!
still we wait for next generation M Chip! And even this will get bypassed by corrupt authorities!
@evacide I solved my age verification issues by not requiring it / refusing to implement and refusing to comply. Any site that demands it, I just don't use anymore. I'll be back on IRC with dialup before I submit to that nonsense.
@evacide Thanks for posting this. Server side age verification, even with ZKP, is not private. Moreover, with Hardware Attestation, it effectively signs the death of Open Source:
Changing a single line of code will invalidate your attestation and ban you from society
Local parental-like control is the best solution: more robust, can't be bypassed by VPN, can block uncooperative websites (add their IP/domain to the local block list) & compatible with competition, free society and open source
@evacide
Couldn't having to read an analog clock suffice?
@evacide Thank you for posting this. I was guilty of thinking ZKP *might* be a solution, but this reality check has changed my mind firmly to the negative.
@TallSimon I'm glad to hear it.
@evacide From a technical perspective, if zero-knowledge proofs work like authentication “tokens,” then there’s a scale and complexity problem. Implementing authentication tokens within a single organizational network is difficult enough, what with all the different kinds of users, accounts, platforms, and applications it needs to work on. But…people think this kind of thing is workable for the whole Internet?
@evacide Classic case of technical solutions trying to solve people problems. Just doesn't work. At least not without ridiculous costs