One thing I’ve noticed over the years is that in terms of marketing, reddit has a disproportionately high level of return in interaction relative to its size, while Twitter has traditionally had a low level of return relative to its size.

For some reason, comments on reddit has always been viewed as more trustworthy relative to other social media platform, despite reddit or’s general reputation for being confidently incorrect on many subjects.

There are certain people whose entire career was made by their reddit posts, yet, it was always odd to me that reddit never managed to effectively capitalize on this other than making their platform worse with every update.

Testing out this theory has been interesting.

This is obviously a deepfake if I ever seen one: if this is a really a visitor centre, then why isn’t it located in the centre of the crater?

You probably never heard of it, but I shitpost on this obscure technology forum called Lemonworld or something on things I would never admit to in public.

It’s great, I’m something of a celebrity over there.

Some of us are 30 year old women, than you very much.

The Joker.

I really don’t get the obsession with overusing the Joker when most of Barman’s other rogue gallery members are much more interesting. The worst of it is during “The Batman Who Laughs”, because now you have both the regular Joker and “what if Batman became Joker” running around. And then, as if it couldn’t get any more ridiculous,

And I think most people will agree by now that Harley Quinn is a lot better of a character after she’s done being Joker’s sidekick.

🥰

That’s the reason most big name actors have their own production companies to work on their own movies, you have way more control over your own role compared to being just another actor on set.

In any case, I think the audience still recognize great performances by individual actors even if the end product is badly reviewed or didn’t do well at the box office, and there are lessons to be learned from the “Suicide Squad” and “Amsterdam” and “Babylon” until you finally get a “Barbie” against all expectations.

I hope he finds his own “Barbie” too.

I just play her on TV.

Yay

Maybe she’s busy working and can’t spend as much time messing with tech friends on social media or something.

It is clearly July 21st, as shown on the banner of this community.

I mean, there probably are celebs out there, but under pseudonyms.

Hypothetically speaking, if a famous person did run around Lemmy under their real name and openly promoting their work here, who is going to believe them? 🙂

Way ahead of you on that.

There is always risks in going to a new social platform, but also the reward for being an early adopter is also great if the platform gets big, and I’d say the Barbie promotion here last year overall went pretty well.

And it comes in pink!

(Magenta, but still)

On the other hand, all these AI errors by Google have made for some great memes recently.

Windows 11. Because Windows 7 ate Windows 9.

Yay

That’s esteemed Academy Award nominated character actress (and producer) Margot Robbie to you!

I knew it! So that’s what you’ve really been up to on Lemmy, @kjaeselrek@lemmy.ml

Or should I say, Academy Award nominated actor Woody Harrelson?

Reddit, and by extension, Lemmy, offers the ideal format for LLM datasets: human generated conversational comments, which, unlike traditional forums, are organized in a branched nested format and scored with votes in the same way that LLM reward models are built.

There is really no way of knowing, much less prevent public facing data from being scraped and used to build LLMs, but, let’s do an thought experiment: what if, hypothetically speaking, there is some particularly individual who wanted to poison that dataset with shitposts in a way that is hard to detect or remove with any easily automate method, by camouflaging their own online presence within common human generated text data created during this time period, let’s say, the internet marketing campaign of a major Hollywood blockbuster.

Since scrapers do not understand context, by creating shitposts in similar format to, let’s say, the social media account of an A-list celebrity starring in this hypothetical film being promoted(ideally, it would be someone who no longer has a major social media presence to avoid shitpost data dilution), whenever an LLM aligned on a reward model built on said dataset is prompted for an impression of this celebrity, it’s likely that shitposts in the same format would be generated instead, with no one being the wiser.

That would be pretty funny.

Again, this is entirely hypothetical, of course.