Hacker News Reader

Comment by tmoertel | Hacker News Reader

I'd rather have the ability to buy a Google Bond for $X that Google must return to me in the event they lock or terminate my account. That way, if their algorithmic classifiers on a whim decide that I'm a bad person and they should lock my account, Google would have a financial incentive to invest up to $X verify that I wasn't the victim of a false positive before pulling the rug out from under me.

View on HN

For a company with $60,000,000,000 in annual revenue, what should the price of the bond be?

What amount would make them think investigating your lock is worth the effort?

It's not so much that they would be desperate to keep the $x; but rather "has had $x posted for Y years" is a strong signal the account warrants careful human review for any flagged issue.

Surely any actual nefarious user would see that $x as a small cost of doing business...

Why do you think that? What kind of nefarious use makes a hundred bucks or more per account and is subtle enough for the ban review to let them keep doing it?

There is no cost to a bad actor in the OP's proposal. If/when the account gets terminated, the money is returned to them. This policy would be worse than useless. No real users would sign up for it, only abusers.

Well the bond will be meaningless if it was only put there yesterday. If it's been there for a while that's a real cost. But true, it's not the full amount unless they do something that makes them owe Google money.

For bad actors, an extra careful review is still going to result in account termination. So, for them, the bonds buy virtually no protection, but it does tie up real money. How many bad actors are going to prefer to loan Google $100 per active account rather than $0 per active account?

Maybe $1000 is reasonable amount both for me and GOOG. BTW they might just not to give you anything anyway, for example if some algo considers you a terrorist who is strictly forbidden to be funded.

$1,000 is their daily budget for M&Ms. You need to go a few orders of magnitude larger.

Across the entire user base is a few orders of magnitude larger.

But they'd never block the entire user base, right? So it's only about a few one offs. M&M money, no?

0.5% of the entire userbase (which might be a fair estimate of their false-positive ratio) is a sizeable number.

Buy some adsense revenue and never cash it out. You would have to be an extra bad person before they consider banning you. If they do it's still dubious to terminate a business relationship while you owe the other person money.

They would kinda not want to go in front of a judge even if he states they are allowed to seize peoples funds without any explanation.

Google will happily terminate an AdSense account and confiscate the earnings. Has happened for years, including to prominent sites like BoingBoing and Fark (https://www.inc.com/erik-sherman/you-wont-believe-the-reason...).

But they will state a reason for doing so. The "incentive to verify" was there. Ofc it doesn't change that it is an evil corp that (if at all possible) you shouldn't get involved with.

LOL, no they won't.

https://support.google.com/adsense/answer/57153?hl=en

> Because we have a need to protect our proprietary detection system, we're unable to provide publishers with any information about their account activity, including any pages, users, or third-party services that might have been involved.

The trial by robot really cracks me up. How is this even legal?

I'm still convinced there is some small / non zero difference between banning free users and banning users that produce revenue but it might just be as insignificant as you make it out to be.

Reply on news.ycombinator.com