41 ms·
In my opinion a simple regulations would target notifications and feeds: 1. You can't enable content notifications by default. (unless it's system or performan
by 627467 6y ago
In my opinion a simple regulations would target notifications and feeds:
1. You can't enable content notifications by default. (unless it's system or performance related one) users have to manually turn them on
2. You can't nudge users to enable notifications.
3. You have to enable switching off notifications easily
4. By default feeds have to be strictly time-based. No 'recommendation engines' by default. No exceptions.
5. Related to 4. No paid posts. Advertiser can buy banner ads, interstitial, but those should not be presented as posts. Should be clearly marked and displayed distinctively from feed content (content I get because it specifically looked for and followed)
I think the points above will take us back to social media of 10 years or so ago. But if it not suficien, here's another one:
6. Platform cannot recommend accounts to be followed. Follows need to be discovered actively by user
- rusk 6y ago> 4. By default feeds have to be strictly time-based. No 'recommendation engines' by default. No exceptions Yes! Only recently twitter have reenabled this “feature” - I can now safely browse without some algorithm showing me things that enrage me. It’s a much nicer place to be now. EDIT I’m quite happy wrt (5) how twitter to the promoted adds. They’re clearly labelled and easy to block if I don’t like them. The algorithm soon learns what to show if you take a zero tolerance approach.
- lemming 6y agoWhat about: 0. Platforms may not customise information presented to an individual user based on knowledge the platform has about that user. I guess the issue with that is how far do you go - a webmail client could be considered to be customising what each user sees, because each sees different email. However it's the level of personalisation which is the root of the problem. Perhaps instead: 0.b) Platforms may not customise advertising material presented to an individual user based on knowledge the platform has about that user.