Margot Robbie
On the internet, nobody knows you are Australian.
also https://lemm.ee/u/MargotRobbie
To tell you the truth, I don’t know who I am either. Somebody sincere, perhaps.
But if you ever read this one day, I hope that you are as proud of me, as I am of the person I imagined you to be.
- 2 Posts
- 86 Comments
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.
The precedent in this case already exists in Midler v. Ford Motor Co., in which when Academy Award nominated actress and singer Bette Midler sued Ford after Ford hired musical impersonators to sing famous songs for their commercials.
The court ultimately ruled in favor of Midler, because it was found that Ford gave clear instructions to the impersonating actress to sound as much like Midler as possible, and the ruling was voices, although not copyrightable, still constitutes their distinct identity and is protected against unauthorized use without permission. (Outside of satire, of course, since I doubt someone like Trump would be above suing people for making fun of him.)
I think Scarlett Johansson has a case here, but it really hinges on whether or not OpenAI actively gave the instruction specifically to impersonate Scarlett’s voice in “Her”, or if they used her voice inside the training data at all, since there is a difference in the “Sky” voice and the voice of Scarlett Johansson.
But then again, what do I know, I’m just here to shitpost and promote “Barbie”.
Most serious artists have switched to using ArtStation and/or Instagram a long time ago, not because of AI, but because of the weird stuff on DeviantArt.
This article also misrepresented Andersen v. Stability AI, you can read the judge’s opinion here:
https://admin.bakerlaw.com/wp-content/uploads/2024/01/ECF-117-Order-on-Motion-to-Dismiss.pdf
But basically, the judge’s opinion was scathing and dismissed all but one of the plaintiffs’ claims.
There is a very good reason you can’t copyright artistic styles.
Wait, you mean to tell me that the beautiful woman that you are talking to online could actually be a marketing team that is creating the idealized image of said woman designed specifically as a commercial product, like some kind of movie about a plastic doll, Barbie, now available on Blu-ray and select streaming services?
Who would do such a thing?
Plenty of companies pivot when their initial business plan doesn’t work out or if the market changes.
For example, did you know that American Express used to be a shipping company like FedEx and UPS? Or Mattel used to sell picture frames before they started making toys? Or Nintendo started out as a playing card company and still sells them today?
They should switch the name to Wayland.com as a compromise.
Live by the sword, die by the sword.
AI generated images are not, and should not be considered copyright able, and they don’t own the right to the image they generated, as I understand it.
Otherwise, Midjourney are certainly very welcome to start paying royalties to certain popular celebrities whose images they are profiting off of. You can’t have it both ways.
I will offer an alternate explanation:
Restaurants (and by extension, coffee shop) are inherently very risky businesses to start and operation, I think over 90 percent of them fail in the first year. So, to increase the chances of survival, you have to make sure your business is ran as efficiently as possible, things like rent and location is outside of your control, so you always want to maximize the value of every dollar you spend.
The first people notice when they go into a new restaurant is cleanliness, nobody wants to eat at a new place that looks dirty, you can only get away from that if you are a decades old local hole in the wall. Ease of cleaning, therefore, is the number 1 priority over everything else.
So, why the “industrial” concrete/tiled floor and metal chair? Because you can just hose them down at the end of the day. Same thing with big open wooden tables and sparce renovations, ease of cleaning.
The second thing is you have to avoid major renovations, make the with the space you have and maximize the amount of interesting decorations for the minimum money/work.
Why put subway tiles on white walls for decorations? Because you don’t have to hire painters, and subway tiles are cheap and interesting looking.
Why use big open windows and only dim Edison bulbs for lighting? Because hiring electricians to rewire the place you rent for lighting is a lot more expensive than using the big windows you already have.
Why avocado toast? Because coffee is your main focus, the food is important but secondary, and a piece of fruit on a piece of bread pretty much doesn’t require any cooking.
It’s really the operation efficiency, rather than some trend following “Instagram” asthetics that led to all these coffee shop looking the same, I think this is a better explanation than what this article proposes.
We still have a work account, along with fan pages and memes, etc.
Hoping the Lemmy shitposting meme magic will work again, I don’t understand how it works, and it did backfire during the Golden Globes when your favorite esteemed character actress to you got her own Lemmy bit turned around on her:
Koy continued: “The key moment in Barbie is when she goes from perfect beauty to bad breath, cellulite, and flat feet — or what casting directors call ‘character actor’
It is exactly because Instagram is at the scale that it is that caused moderation to be difficult. Facebook has relied on using bots to moderate for so long due to its scale, and using bots that are specifically designed to detect AI generated contents is really not possible without introducing a ton of false positives, since the Instagram of the 2020s at its core IS celebrity/influencer advertisement, and there is honestly very little that differentiate what constitutes as "content* and “spam” there.
Since influencers will be the first to be automated by machines, I just don’t really see a point in having an Instagram account any longer, the inevitable conclusion of creating a fake reality of your life on Instagram is being replaced by a machine that can fake it more efficiently.
So, this week, we are talking about deGoogling on !android@lemmy.world.
https://lemmy.world/post/9772354
You guys are very welcome to join in the conversation.
Fellow lemmings, I, for one, am glad that there are no corporate trolls trying to manipulate public opinions on Lemmy, it is the same warm, fuzzy feeling I get when I was watching multiple time Golden Globe nominated summer blockbuster, Barbie, now available on Blu-Ray and select streaming services.
I don’t know about you, but I sure hope Barbie sweeps the Golden Globes next month (and then the Oscars next year.)
Replacing the buttons with a tablet has always been a cost saving measure on Tesla’s part that was marketed as “futuristic”, physical switches and dials made of plastic and metal as well as the underlying components will never be as cheap or as easy to wire as a simple touchscreen control. Other car companies followed suit, because Tesla made a method of reducing their own manufacturing costs hip, so many of them jumped on it.
But, Tesla tablets were designed with the belief that this cost saving is possible because of the delusion that full autonomous self driving is possible with existing hardware through software updates. When self driving didn’t happen after a decade of trying, people realized how inconvenient and dangerous it is that the only way to adjust the AC, stereo volume, and sideview mirrors while driving is through a tablet with no tactile feedback. So now, we are finally seeing that trend reversing.
One of the great thing about the AI revolution is that since generating infinite number of unique random (and commonly, bad) pictures of literally anything you can think of takes only seconds, the entire concept of NFT has become completely worthless as it completely destroyed the value-from-scarcity argument. Not that it ever was a good argument to begin with.
Let me address some of your points:
a forum like discourse, which is most definitely not a classic forum.
Linear post structure, sorting based on latest response, so it is a traditional in vein of old BBS/phpBB systems.
Do you really think you could replace the utility of stackoverflow with… Reddit?
Upvote based sorting and nested comment structure means StackExchange/Overflow is closer to reddit than it is Discourse.
It really is not much of a step to take your logic and replace lemmy or Reddit with Discord as a Q&A support and knowledge base platform.
Reddit, Lemmy, and Discourse are all public forums, Discord is a chatroom.
Even simple features that encourage engagement in months or years old threads are massive boon.
Thread necromancy for month/year old dead threads has always been considered offenses to almost every single forum, which is why most forums lock posts after a month or so. It’s not a feature, it’s a fundamental flaw with the sorting.
with copious features built specifically to enable and support that purpose. Which both lemmy and Reddit lack
I’m genuinely curious, what are some of those features? I can’t think of any significant one, outside of tags.
Forums are ultimately shaped by people, so I would say these forum succeeded in spite of the software instead of because of the software.
What do you get when you take a traditional forum, add vote based sorting so that the best post and comments (in theory) rise to the top to avoiding the issue of thread bumping, use a nested comment structure so that individual conversations in each thread can be easily followed, and allow anyone to make a subforum as they please?
You get reddit. (or now, Lemmy) The only thing missing in Lemmy is topic tags, which I think is a nice to have, but by no means necessary.
There is a reason why very few people uses forums and most of conversation nowadays takes place on social media, while reddit and Discord has but all but replaced them. So, replacing Discourse/Zenforo as the software to use for independent Internet forums should be the aim for Lemmy to significantly grow.
Strangely, Discourse’s federation model seemed to be based on Mastodon compatibility instead of Lemmy-like Groups, which I think is a mistake.
Not exactly surprised here. Every time I’ve seen him on the news, it’s always him fearmongering about the dangers of generative AI, when ChatGPT is burning through money and seemed to become more and more restrictive with every iteration. You can’t run an organization if it is built on top of lies.
Actually open models (not open source, sadly) like specialized LLaMa 2 derivatives that could be ran and fine-tuned locally seems to be the future, because there seems to be a diminishing return in training/inference power to usefulness, and specialized smaller model tuned for specific applications are much more flexible than a giant general one that can only be used on somebody else’s machine.
I don’t know, maybe he’s a big fan of the Barbie movie (now available on streaming on Amazon Prime videos) and is cosplaying as the Patriarchy. Not even a Ken, but like, the literal personification of the Patriarchy.
Cowboy and cowgirl costumes are fun to wear, but they are just that, costumes. The only acceptable reasons to pretend to be someone you’re not on camera is either you get paid for it as your job, or if it is really, really funny.
Twitter’s influence has always been overstated by the media type, and the personalized recommendation algorithm it used has always been a black box to be gamed, and what this article really showed is that it is an inheritly unfair game and the people who writes the rules of the platform (in this case, Musk), can push whatever hot take they want to push to influence public opinion, and people who don’t have that level of access to Twitter will brute force follower with bots in the hopes that it’ll lure more real users to influence.
So, dead internet theory and all, until we fully automate Twitter to bots influencing bots. There is no authenticity anymore, it’s all disinformation from all sides.
Nowadays, I like to pretend that I never had a Twitter or Insta. Not a healthy way of coping, I know, but it’s way less stressful that way.
Why do you think Sam Altman is always using FUD to push for more AI restrictions? He already got his data collection, so he wants to make sure "“Open”"AI is the only game in town and prevent any future competition from obtaining the same amount of data they collected.
Still, I have to give Zuck his credit here, the existence of open models like LLaMa 2 that can be fine-tuned and ran locally has really put a damper on OpenAI’s plans.
This is almost an “arms race” situation, since when there are so many gigantic SUVs and pickup trucks on the road, driving in a smaller car becomes a lot less safe in case of an accidental collision with a larger, heavier vehicle, and the only way to reduce that risk is to drive a gigantic SUV/pickup truck yourself and further exacerbate the problem.
Having this many large vehicles on the street makes driving on the highway dangerous and unpleasant. LA’s traffic is especially terrible.
It’s made by Ben Zhao? You mean the “anti AI plagerism” UChicago professor who illegally stole GPLv3 code from an open source program called DiffusionBee for his proprietary Glaze software (reddit link), and when pressed, only released the code for the “front end” while still being in violation of GPL?
The Glaze tool that promised to be invisible to the naked eyes, but contained obvious AI generated artifacts? The same Glaze that reddit defeated in like a day after release?
Don’t take anything this grifter says seriously, I’m surprised he hasn’t been suspended for academic integrity violation yet.
And Netflix and other streaming services, despite supposedly saving us from cable, re-invented cable.
But the good news is, maybe it’s time to get the good old iPods/DAPs back into the mainstream again. You can have a big SD card with all the music and podcasts you could ever want without tying yourself to a specific service.
Which goes to show that if you are a public figure whether in politics, entertainment, or otherwise, owning your own server for social media instead of relying on Zuck and Musk should be a critical concern at this point.
There is no reason for the Harris campaign to not dip their toe into federated social media at this point given Musk’s antagonism towards them.