YouTube AI deletes war crimes evidence as 'extremist material'

Agree. If Google just wants Youtube to be an entertainment funnel, it has every right to do so.

If the service you describe exists, it won't survive long. People would begin to swarm such services with rejected videos from Youtube, which you can foresee will be problematic and messy, in the end it might just turned into a 4chan where you can upload video. Hardly an attractive business.

I think the problem is the expectation of people that someone else do the filtering for them. I.e. "I don't want to see this kind of content" leads to "someone else should remove it from all the sites I visit". Which obviously leads to conflicting requirements once you have more than one person and those people disagree on what they want to see and don't want to see.

The only reasonable solution is to host everything, modulo requirements by law, and give users the tools to locally filter out content en masse.

In a decentralized system you also skip the law requirements since you cannot enforce multiple incompatible jurisdictions at the platform level, individual users will be responsible for enforcing it on their own nodes, similar how all you can do when accidentally encountering child porn is to clear your cache.

The problem on these distributed platforms is not filtering what people see, but filtering what people host or allow to transit their network connections.

> In a decentralized system you also skip the law requirements since you cannot enforce multiple incompatible jurisdictions at the platform level, individual users will be responsible for enforcing it on their own nodes, similar how all you can do when accidentally encountering child porn is to clear your cache.

But that's the thing: You don't skip it. You spread it to every user. They both have to deal with whether or not they are willing to host the material and whether or not it is even legal for them.

How many of us sympathise with the idea of running a Tor exit node, for example, but avoid it because we're worried about the consequences?

These platforms will always struggle with this unless they provide ways for people to feel secure that the content that is hosted on their machines is content they don't find too offensive, and/or that traffic that transit their networks is not content they find too offensive.

Consider e.g. darknet efforts like cjdns which are basically worthless because their solution to this was to require people find "neighbours" they can convince to let them connect. Which basically opens the door to campaigns to have groups you disapprove of disconnected by harassing their neighbours and their neighbours neighbours, just the same as you can go to network providers on the "open" internet.

> That's not a free market though.

> markets take the path of least resistance to greatest profit

I assert that a society in which love and compassion are the orders of the day will, in the demands and price signals it sends to the market, not abide by this orthodoxy.

> if you want "societal good" to be guaranteed, (given some definition of that) you have to force it to be so - which is what the violence of the state is intended for

The idea that wonderful ends can spring from such unseemly means is not in keeping with what I am able to observe of the universe.

First of all, not all p2p networks operate like Tor. For example bittorrent and ipfs only host content you look at. So hosters could largely self-select the content they replicate.

Secondly, there are several tiers of content. a) stuff that is illegal to host b) stuff that is not illegal but that you find so objectionable that you don't even want to host it c) stuff that you don't like but doesn't bother you too much d) stuff you actually want to look at. I posit that a) and b) are fairly small fractions and the self-selection mechanism of "things that I looked at" will reduce that fraction even further.

And even if you are on a network where you randomly host content you never looked at encryption can provide you some peace of mind (of the obliviousness kind) because you cannot possibly know or expected to know what content you're hosting. Add onion routing and the person who hosts something can't even be identified. If Viewer A requests something (blinded) through Relay B from Hoster C then B cannot know what they're forwarding and C cannot know what they're hosting. If neither you nor others can know what flows through or is stored on your node it would be difficult to mount pressure against anyone to disconnect.

For the illegal content, especially in oppressive environments, you could install a Voluntary Compliance(tm) government blocklist on public-facing nodes and still opt to run an internal node in your network that uses encrypted connections to retrieve things hosted in other countries you're not supposed to see.

----

Anyway, back to filtering for decentralized content hosting. I think once you have a network it is a matter of managing expectations. You can't make content magically disappear. Platforms like youtube, twitter, facebook etc. have raised the false expectation that you can actually make things go away by appealing to The Authority and it will be forever gone. In reality things continue to exist, they just move into some more remote corners of the net. Once expectations become more aligned with reality again and people know they can only avoid looking at content but not make it non-existent things boil down to being able to filter things out at the local level.