Yep the flaw is assuming that humans would actually select for constructive comments. It’s a case where humans claim that’s what they want, but human actions do not reflect this. We’d eventually build yet another ‘algorithm that picks what immediately appeals to most users’ rather than ‘constructive’. You’d also see the algorithm splinter along ideological lines as people tend to view even constructive comments from ideologies they disagree with unfavorably

They just need to pull shorts out into it’s own app. It’s not very often that I’d want to freely mix short 30 second videos in between longer YouTube content. They’re different use cases.

The only issue with your second point is that it can eventually become a quagmire when you do need to upgrade it.

I work for a very old company who held to that philosophy for many years. And while any individual component could be looked at and seen as running fine, when they did finally decide it was time to upgrade they were faced with needing to upgrade everything simultaneously.

All of the tech was too old, so no current tech had the sort of backwards compatible bridge that helps you move forward. It’s like figuring out how to get your telegram system to also work on your WiFi network, nobody makes any interfaces for that.

Instead of slowly and gradually replacing components over time, they’re faced with a single major overhaul that’s put the entire company at risk because they have to completely shut down for over a month.

Have to setup port forwarding to make it available outside your network.

I tried running it off my main PC for awhile, but it’s kind of a pain and impacted my gaming sessions. Moved it onto a $400 Lenovo server I got off eBay to great effect. Used unRAID to setup a simple server for it. Lots and lots of guides out there for it with a great community. Have since upgraded beyond that server, but it was a great start.

Setting up a world in which you are forced to drive and then making incredibly draconian surveillance of your performance of that required task is just cruel. Put this effort into providing me travel options that don’t come with the risk of major injury, death or jail time.

Until TV is setup the same way Spotify/YouTube music/apple music is where you just pick one you like and listen to the same music the other platforms have, they’ll continue to have pirating problems.

I currently pay more per month for the various components needed for highly effective pirating than I would for cable and that’s purely because it offers a better experience. I can’t buy a plex-like experience anywhere for any price legally.

Fix that and I’ll go legit just like I did for music.

Most likely the entire HBO streaming service wouldn’t have taken off, because they offered little to no avenues to consume their content to an increasingly no-cable subscription generation. It’s entirely likely that HBO would’ve died out along with traditional TV.

Just since I’ve setup a plex server (about 8 years now) midrange sizes have gone from 4->16 TBs. Personally I think the bulk of the issue is that HDD customers switched from a mix of enterprise and personal, to nearly all enterprise. Companies really don’t care if a HDD is $200 or $500, so basically all high capacity drives are priced at B2B prices, not consumer

Struggling to sort out my thoughts on this one.

I’m not really sure comparing AI to a human artist learning and being inspired by others quite fits. At least in the context of a commercial AI (one that a company charges others to use). It feels scummy for a company (for profit entity) to steal training data from others without consent, and then turn around and charge people for the product they built on that stolen content.

That said, existing copyright law allows for ‘fair use’, which includes educational purposes. In that light, AI companies could be seen as a sort of AI school program. But the icky part to me, is that AI is not a person. It can’t choose to leave the school. That school can then profit off that student forever and ever.

I feel like the fair use argument for education applies to humans, not AI (at least not till they actually gain sapience). AI are machines that can be leveraged and exploited by the few and powerful, and that power should come without us subsidizing their development.

Though honestly it’s sort of a moot point, because it’s already done and we’re very unlikely to ever properly charge them now. And now that they have the start, they have a leg up on everyone else. So the morality of how it was built no longer really matters, unless we want to argue AI should all be open source or public domain.

We shouldn’t have to pay for the content we use to train and teach an AI

Wait people think that sounds reasonable?

Yep. The Fediverse has a lot of growing room in the QOL department and is hampered by the relatively small (and often part time) dev teams working on it. Meta comes in, builds a compatible platform, then starts offering meta-platform only ‘improvements’ that offer those QOL features. Rest of the Fediverse dies out because ‘meta’ isn’t that bad and they aren’t abusing their position (yet).

This. I’m happy with these content offerings, but it’s weird to have them all mashed together. It’d be like if Lemmy randomly inserted book chapters in between other posts. It just doesn’t fit with the platform and the use cases are wildly different.

Most of my spam is already robots. But I can see the call network going down because all calls are now just robots talking to robots with no exit loop. Would be the final nail in the coffin for voice calls for me.