Gmail spam filtering? Isn't particularly good. I run my own email infrastructure and I get way less spam through it than on my gmail address. And I don't do anything particularly special other than configure postfix correctly.
Recaptcha? Was there ever a more hated piece of user abusive technology? Plenty of rage against it here in many discussions. I won't do business anymore with any site that won't let me buy without identifying 27 traffic lights in slow motion (looking at you adorama). We're all better off if this dies.
Facebook verification? Don't know what value this brings to anyone other than FB.
Youtube? Granted.. Hopefully with bandwidth costs reducing over time, decentralization becomes more viable.
Paypal? Always pay with credit card directly. The regulatory consumer protection for that was well in place before internet commerce even took off.
The original potential of the internet was in its fully decentralized functionality. Usenet being a great example (owned by nobody, hosted peer to peer) which should be studied a bit more now that we have generations of developers coming up never having known it. Of course, it was (mostly) overrun by spam. But unlike email where spam is pretty much defeated, there was never a meaningful effort in addressing spam on Usenet (forums and mailing lists took over instead). Applying the same techniques to NNTP could well be equally successful.
If there is any desire to avoid the internet becoming fully a new cable-tv controlled by just a few major player we need to pursue research into decentralized services as a priority.
(If nothing else, consider at least running your own email infrastructure if you're technically capable. Keep the foot in the door so it doesn't close permanently.)
And while I hate using recaptcha, I also had the even more unpleasant experience of watching a forum that I moderated die under a crushing assault of spam bots that were smart enough to make it past the built-in CAPTCHA. Recaptcha isn't there to solve problems for the people who interact with it, it's solving problems for the people who chose to put it on their site.
I wrote in a sibling comment what value Facebook's login brings to the sites that opt into it. (Offload all the work of password handling, claims about stolen accounts, preventing sockpuppetry and catfishing, etc).
But there should be accessibility laws to prevent government agencies from blocking access to citizens/taxpayers using these proprietary technologies.
I don't know offhand of an open-source captcha library that's nearly as effective as reCAPTCHA. All the old "wavy lines through slanted text" ones are completely broken with current OCR techniques.
> Facebook's identity verification
in real life, do you need to know the real name of the cashier and see their contract so you can trust them with your payment? You don't, you use common sense.
> Google de-ranking link farms
hey, haha, and what do we get instead? Monopoly on ad-driven search engine, great, I'd rather have link farms since I can easily filter those out with little brain-power. I can't do that with paid-for-propaganda
Now imagine we'd leave the way it was, people would learn how to deal with it, the same way you learned you can't leave unlocked bicycle in a city. It's a price to pay for anonymity, and we should all have right to be anonymous on internet.
A couple of ideas I've had:
- Leverage PGP and trust rings for identity without a need for non-anonymity. This could be structured in a nuanced manner so that we respect pseudo-anonymity while also maintaining the integrity that real identity brings.
- A wholesale education of the populace, instead of an education by fear, of the power of anonymity and services which further anonymity.
- A move away from the advertising and services-exploitation business model. Not that we can't have service oriented businesses, but those businesses cannot be exploiting people to do labor and a bottom line price.
- A data ownership model which establishes a monetary value for different types of data. We already do data classification inside mega-corps, and we're able to establish and correlate it's sensitivity. Extrapolating this to value would not be stretch. Users consent which data they give up in exchange for services. The price for the service is adjusted based on what data they give up, and how they choose it will be used (Data for marketing may be more valuable than say data for behavior research, or even vice versa!)
> Facebook's identity verification.
The ability to choose an identity online not connected to your legal or in-person name, to make it more difficult to find and stalk you in person, is a feature, not a bug.
For these purposes, an actual Real Verified Name is overkill, and a unique identity number would be enough. But Facebook offers that service for Free(TM) so there's not much incentive to the site operator to aim for something more minimal. And with GPT-3 and other systems inching closer to passing the Turing test, this only makes alternatives more challenging for them.
Those are problems in the Wild West that would need solutions if we were to reclaim the net from Facebook and Google.
(Google no longer rejects or bans you for fake-sounding names, ever since the Google+ controversy, but it still enforces a Western-style first/last name instead of other cultures' name formats or single-part usernames.)
Gmail's spam filtering. Recaptcha's bot rejection. Facebook's identity verification. Paypal's secure credit card processing. Google de-ranking link farms. Youtube's hosting costs. All of these very effectively solved real problems with the old web, and those problems were growing as bad actors became more sophisticated.
I like the old web too, but I do appreciate that an attempt to return to it would still require very effective tools to solve at least some of these problems, and they'd require constant improvement to keep the roving bandits at bay.