Live data from Hacker News

Google, Meta, others will have to explain algorithms under new EU legislation

theverge.com

51–60 of 222 posts

Re: Google, Meta, others will have to explain algorithms under new EU legislation

#51
post #22
post #16

Earlier quoted context omitted.

> Google doesn't know what the algorithm is anymore You are an insider?

They haven't talked much detail since Matt Cutts left, but over time they did sort of outline the basics. That the core ranking is still some evolution of PageRank, weighting scoring of page attributes/metadata and flowing it down/through inbound links as well. But then altered via various waves of ML, like Vince (authority/brand power), Panda (inbound link quality), Penguin (content quality), and many others that ta…

Those sound like awesome potential features. Allow users to assign 0-100% weights for each of those scoring adjustments during search,and show them the calcs (if you can).

Re: Google, Meta, others will have to explain algorithms under new EU legislation

#52
post #40

A good start. However let's go further, simply ban personal tracking and personalized algorithmic feeds. This would combat the echo chamber effect and social media can become a broad community experience, like TV and newspapers. It would also cripple tech advertising revenues, thus redressing the balance with traditional media.

I’ve seen the difference between what YouTube presents to me when I’m logged in v.s. when I’m on a clean computer it can’t associate with me, and I do value the personalisation — when not logged in it shows me a hundred duds for every one thing I care about, and logged in it’s about 50:50. How much of this improvement is a mysterious machine learning algorithm and how much is it just looking for new things from my su…

An alternative might be for personalisation to be opt-in rather than opt-out even when signed in with an account (which shouldn't even be necessary for many services anyway)

Re: Google, Meta, others will have to explain algorithms under new EU legislation

#54
post #39

> The greater the size, the greater the responsibilities of online platforms > as a rule, cancelling subscriptions should be as easy as signing up for them Overall I like these principles, but we'll see in a few years how they're enforced in practice. It's been 4-5 years since we've had GDPR and I still see sites that require tens of clicks to disable all advertising cookies (and the most I've seen was 300+ clicks).…

4-5 years is nothing for the law. You have murders from 15 years ago still in process in courts. But eventually things settle. It just takes time. It's a bit like ents

Re: Google, Meta, others will have to explain algorithms under new EU legislation

#55
post #46

Earlier quoted context omitted.

Amazon's attempts to get users to unwittingly sign up to Prime is one of the most egregious examples I encounter on a regular basis. As a European I cannot wait to see it gone.

I feel like I remember it being pretty easy to cancel Prime, though. Have things changed?

Yes: https://www.msn.com/en-us/autos/news/yes-it-seems-amazon-did...

Re: Google, Meta, others will have to explain algorithms under new EU legislation

#56

A good start. However let's go further, simply ban personal tracking and personalized algorithmic feeds. This would combat the echo chamber effect and social media can become a broad community experience, like TV and newspapers. It would also cripple tech advertising revenues, thus redressing the balance with traditional media.

I think it's fair to use personal data collected on the same site. Without it, most sites would be rendered useless.

Re: Google, Meta, others will have to explain algorithms under new EU legislation

#57
post #46

Earlier quoted context omitted.

Amazon's attempts to get users to unwittingly sign up to Prime is one of the most egregious examples I encounter on a regular basis. As a European I cannot wait to see it gone.

I feel like I remember it being pretty easy to cancel Prime, though. Have things changed?

Doesn't excuse tricking people into it.

Re: Google, Meta, others will have to explain algorithms under new EU legislation

#58

Earlier quoted context omitted.

"I know it when I see it."

in common law systems I know it when I see it is good enough, but I believe most of EU is under a Napoleonic system where you should define what you mean.

The requirement of "subscribe is as easy as unsubscribe" is a metric which you could argue about in court, but would be very hard to game.

i.e. if signup is "email and credit card number" then you're going to be hard pressed to explain why a similar option to cancel does not exist and isn't accessible in as many clicks, with equivalent screen real-estate usage.

Re: Google, Meta, others will have to explain algorithms under new EU legislation

#59
post #46

> “Dark patterns” — confusing or deceptive user interfaces designed to steer users into making certain choices — will be prohibited. The EU says that, as a rule, cancelling subscriptions should be as easy as signing up for them. This is an excellent addition.

Amazon's attempts to get users to unwittingly sign up to Prime is one of the most egregious examples I encounter on a regular basis. As a European I cannot wait to see it gone.

If it's easy to accidentally sign up for something, does that mean it has to be easy to accidentally cancel something? Because that would be hilarious.

Re: Google, Meta, others will have to explain algorithms under new EU legislation

#60
post #40

A good start. However let's go further, simply ban personal tracking and personalized algorithmic feeds. This would combat the echo chamber effect and social media can become a broad community experience, like TV and newspapers. It would also cripple tech advertising revenues, thus redressing the balance with traditional media.

I’ve seen the difference between what YouTube presents to me when I’m logged in v.s. when I’m on a clean computer it can’t associate with me, and I do value the personalisation — when not logged in it shows me a hundred duds for every one thing I care about, and logged in it’s about 50:50. How much of this improvement is a mysterious machine learning algorithm and how much is it just looking for new things from my su…

> I’m not sure, and that’s important: being trapped in a torrent of self-reinforcing falsehoods is something I fell for in my teenage-goth-New-Age phase, which Carl Sagen condemned in The Demon-Haunted World.

Genuinely curious here: how can you tell you've escaped one set of self-reinforcing falsehoods while being sure you haven't fallen into another, different set?

Post reply on HN