muntedcrocodile

It would be a great first filter for calls for violence. And u would give people opportunity for human review with failure costing a permaban.

I know people love to hate on ai but I think that might be the moderation tool needed. Make a bot that just liberally hands out bans and call it Stalin bot. The tankies would love getting banned by that.

If u do pls publish the code under a foss licence and send me a link.

So what ones did I get wrong please tell me.

It was never a fact not prior to or post me googling it. As I said objective reality remains the same regardless of proof.

I know almost everything of ur ideology. U think trump is a fascist along with all his voters. You think dei is a good thing. You support gay and trans rights. U claim to believe in free speech but think misinformation should be censored. You are pro choice and think that people should have bodily autonomy except in a pandemic they must wear a mask as u think its causing harm.

Am I wrong?

Well I just googled it and turns out the moon isn't made of cheese. If u did the same for my claim u would find that I am in fact right.

The place u might find some interesting insights are the timestamps of when things get watched. If u got a health tracker I recon u might find interesting correlations between that and watch data.

How will u classify a taste? U can always eyeball it but perhaps embedding it within a vector dB might work. If u also dump in a large catalogue of just general population music sample u can get a full genre landscape and see how u fit into it.

I would probably try implement Netflix's algorithm which is essentially finding what u haven't seen that has been seen by people who watch things similar to what u watch. Essentially just heigh dimensional embedding and some clustering. So u would want at least 1000 (absolute bare minimum) users.

Data is only really usefull in the context of all other users. So 3-4 users of data on their own don't really give any useful insights.

Recommendation algorithm/engine for self hosted media server. If someone knows where I can get a large amount of watch data for a large amount of people I could write such a thing.

Go Google it. Its not an opinion its fact. As I said not my job to educate you your capable of googling urself.

Didn't realise math was woke that's news to me. I do love how people of ur ideology love to use logical abstract arguments when it suits you but say I don't care about people and have no empathy when I do it.

Facts are still facts regardless of if I've provided proof. That's how objective reality works.

Is that based off my summaries specifically or ai in general? I'm doing something pretty unusual its not just piping to an LLM its using a purpose built transformer that takes a set chunk size and summarises it. It then does that for however many chunks required to make up the whole article.

Its essentially a text -> semantic embedding -> text model more simmilar to an image classifier than a standard text model.

That's how I avoid many of the traditional LLM errors and bias. It also has weird failure cases such as its internal embedding of Vladimir Putin is the same as Vladimir Zelenasky.

I can run a trial where it posts both and ask people to vote for what they prefer.

That's actually a good idea whats world limit would u recommend?