I'm just wondering whether there's evidence to support the claim that training of the LES by eating upside down is possible. To me it doesn't seem entirely plausible, but I'm a layman.
What you call an algorithm here is a recommendation engine. I don't see why they should avoid having something like that. Ideally they would have a modular system, so you could plug in your own favorite third-party engine.
It has some interesting discussion, although it also shows how US-centric Lemmy is. Much of the EU has understood why nuclear energy is inherently incompatible with renewable energy and has therefore rightfullly dismissed it.
This comment for example, after about a week or two most of the visibility and interaction of it will drop to zero. At that point, this comment should expire and no longer exist.
That's an incredible naive and egoistic take. Think about all the knowledge that is getting lost by applying this approach. How many times have you searched for some obscure thing and found the answer only on some five years old reddit post? That information would be lost for ever if you had your way.
And to use it with a similar feature set, everyone is using different extensions which also have to be supported by the clients. I know there is this one server implementation (name escapes me at the moment) and Conversations on the client side, but it's hardly the standard and we're not really talking about plain XMPP then anymore.
I'm just wondering whether there's evidence to support the claim that training of the LES by eating upside down is possible. To me it doesn't seem entirely plausible, but I'm a layman.