That’s a valid concern.

It may be hard and take time, but it sounds like it would be worth at least starting the process and slowly take steps to move. If they’ve done this, what’s the likelihood that they’ll do something even worse later?

I’m not sure.

Perhaps getting computer vision that’s reliably better than humans costs a lot of money that a project like OpenStreetMap doesn’t have?

If it helps, here’s a (partial) list of ActivityPub software - I’m not sure why it doesn’t include things like Friendica or Owncast: https://github.com/BasixKOR/awesome-activitypub

But regarding your question, the first example that comes to mind is PeerTube. Not only does it look to me like it was designed from the start with federation in mind (I don’t know this for a fact though), but it also seems pretty innovative with its use of peer-to-peer video streaming. This 2 minute animated video does a good job of explaining what it does: https://framatube.org/w/217eefeb-883d-45be-b7fc-a788ad8507d3

Owncast seems somewhat similar.

It seems that most Fediverse/ActivityPub software is a “twist” on something that existed previously, but there is still a lot of innovation going on, instead of pure copies of existing centralized platforms.

I only read the text, didn’t watch the video, but from the text it didn’t sound like that’s what they were describing. It sounded like they will present images to people and ask them to confirm whether or not there’s an object there?

We introduce “MapTCHA”, a CAPTCHA that leverages the uncertainty of interpreting imagery with computer vision, and provides human verification for AI predictions: users are asked to identify images containing correctly interpreted objects, e.g. building outlines.

Edit: also, here’s the github they seem to be working from - https://github.com/ciupava/maptcha_dev

Edit 2: I hope they succeed, because it would be great to have an open source captcha that benefits everyone.

deleted by creator

Absolutely! What’s weird is that Teslas have been top-rated for crash-worthiness in the past, so there are a few possibilities I can think of:

• They need to be top crash-worthy, because of the stupid autopilot trying its best to kill the occupants
• They need to be top crash-worthy, because otherwise any crash at all would result in a fiery death
• The Cybertruck is an outlier and is not as crash-worthy as the previous Teslas
• All of the above

What was that rule of thumb for taking multiple choice tests? If you don’t know the answer, always select “all of the above”?

Cybertruck will have 14.52 fatalities per 100,000 units — far eclipsing the Pinto’s 0.85.

Holy shit, that means the Cybertruck fatality rate is around 17 times higher than the Pinto’s!

420 microplastics?

And George Soros being some kind of secret mastermind is also a right wing fever dream.

The funny thing is that so is the idea of Melon Husk being some kind of mastermind.

Yeah, that’s how I do it too. Although some instances don’t seem to block it, you would have to find out which, and that may be difficult to find.

I totally understand why they would do it. I don’t understand why or how they block the NSFW instances without them appearing in the list of blocked instances.

lemmynsfw appears to be blocked secretly on lemm.ee. It’s not listed under blocked instances there, but I can’t see any posts from lemmynsfw there either. There must be some other mechanism that’s available besides blocking instances in the open.

Weird. Then either our instances are secretly blocking it, not through defederation, or there’s some other mysterious technical reason for it.

But also think that the rule of Lemmy is that all algorithms end at porn.

Huh? I don’t see any porn, regardless of what sort I choose. Ah, that’s because there’s a checkbox in your settings to let you choose whether to see NSFW posts or not. I guess turning that off works really well.

You have really good insights about this, and it’s a great question. I wish it weren’t buried 5 levels deep in this thread, where few people may see it. Maybe you could write it up as a top level post on the Fediverse community?

I agree that the Facebook format is fine. The real problem with them is the algorithms. Oh, and all the data harvesting, advertising, and now bots pretending to be real people? I guess everything else about Facebook is terrible except the format!

Unfortunately, I don’t know. I read that mobile apps for it are few and far between? There may only be one or no app for Friendica on each platform. I think I also read that there are a couple of apps in beta at the moment?

I heard that it was one of the earliest Fediverse projects. It just hadn’t surged in popularity like Mastodon and Lemmy had after the Twitter and Reddit exodus.

Ah, thank you! That’s what I was actually thinking of, but then I thought I was mistaken.