516
submitted 7 months ago by L4s to c/technology

TikTok ran a deepfake ad of an AI MrBeast hawking iPhones for $2 — and it's the 'tip of the iceberg'::As AI spreads, it brings new challenges for influencers like MrBeast and platforms like TikTok aiming to police unauthorized advertising.

top 50 comments
sorted by: hot top controversial new old
[-] [email protected] 213 points 7 months ago

Everyone with a brain has been saying this would happen for the last decade, and yet there was no legislation put in place to target this behavior

Why does every law need to be reactionary? Why can't we see a situation developing and get ahead of it by legislating the very obvious things it can be used for?

[-] camr_on 61 points 7 months ago* (last edited 7 months ago)

How about a real answer:

All but a few of our legislators have any idea how technology/Internet works. Anything about the Internet that is obvious to the crowd on lemmy will probably never cross the radar of a geriatric legislator who never needs to even write their own emails bc an aide will do it

[-] [email protected] 48 points 7 months ago

So, the first reason is that the law likely already covers most cases where someone is using deepfakes. Using it to sell a product? Fraud. Using it to scam someone? Fraud. Using it to make the person say something they didn’t? Likely falls into libel.

The second reason is that the current legislation doesn’t even understand how the internet works, is likely amazed by the fact that cell phones exist without the use of magic, and half of them likely have dementia. Good luck getting them to even properly understand the problem, never mind come up with a solution that isn’t terrible.

[-] [email protected] 8 points 7 months ago

The problem is that realistically this kind of tort law is hilariously difficult to enforce.

Like, 25 years ago we were pirating like mad, and it was illegal! But enforcing it meant suing individual people for piracy, so it was unenforceable.

Then the DMCA was introduced, which defined how platforms were responsible for policing IP crime. Now every platform heavily automates copyright enforcement.

Because there, it was big moneybags who were being harmed.

But somebody trying to empty out everybody's Gramma's chequing account with fraud? Nope, no convenient platform enforcement system for that.

load more comments (4 replies)
[-] [email protected] 7 points 7 months ago

Huh TIL that the average age of the Senate and the House has steadily increased over time: https://www.nbcnews.com/data-graphics/118th-congress-age-third-oldest-1789-rcna64117

load more comments (1 replies)
load more comments (22 replies)
[-] [email protected] 97 points 7 months ago

Simpsons called it:

[-] PocketRocket 70 points 7 months ago

Oh boy. This is all moving very quickly. People already fall for simple SMS scams, I can only imagine just how many more will be falling victim to this trash in months/years to come.

[-] CeeBee 27 points 7 months ago

People have already been falling for scams that "Elon Musk" was promoting. Naturally I'm talking about these crypto schemes run by scammers on YouTube using a deepfake of Musk. It's been happening for about two years now.

[-] [email protected] 19 points 7 months ago* (last edited 7 months ago)

Bill Gates has been giving away his fortune to some lucky email recipients every year now since the days when you had to pay for the internet by the hour.

load more comments (1 replies)
[-] slaacaa 18 points 7 months ago

Just imagine fans getting a facetime call from a Taylor Swift, explaining they won half-price tickets to an exlusive fan event. Then “Taylor” has to drop out to make the other calls, but will leave them a link for the purchase - only valid for 15 minutes, as of course many others are waiting for this opportunity.

[-] [email protected] 64 points 7 months ago* (last edited 7 months ago)

This is the entire basis of using an adblocker like ublock origin. It is purely defensive. You don't know what an advertising (malvertising) network will deliver, and neither does the website you're on (Tiktok, Google, Yahoo, eBay, etc etc etc). With generative AI and video ads and the lack of content checking on the advertising network this will just get worse and worse. I mean, why spend money on preventing this? The targeted ads and user data collection is where the money's at, baby!

Related note, installing uBO on my dad's PC some 8 years ago was far more effective than any kind of virus scanner or whatever. Allowing commerce on the Internet was a mistake. That's the root of all this bullshit, anyway.

[-] Moneo 25 points 7 months ago

Fuck ads in general. I don't care if they are legitimate or not I don't want to be mentally assaulted every time I try to browse a website.

load more comments (1 replies)
[-] [email protected] 6 points 7 months ago

Allowing commerce ~~on the Internet~~ was a mistake. That’s the root of all this bullshit, anyway.

That's more accurate.

load more comments (2 replies)
load more comments (6 replies)
[-] TropicalDingdong 56 points 7 months ago

Bro I still don't know who MrBeast is.

[-] herr 33 points 7 months ago* (last edited 7 months ago)

Currently largest and most successful YouTuber on the platform (by a wide margin), started out by doing challenge videos about himself (24h in ice, that kinda stuff) that he'd invite friends to as the goody sidekicks causing mischief and making his challenges a little harder/more interesting.

These days, his stuff has transformed into a media powerhouse, all of it is still kinda falling into a challenge category. Now with far higher stakes and involving other people in competitions against each other - think "kids vs adults - group with most people still in the game after 5 days wins $500k" - where several days (sometimes months) of filming all gets cut down to one 10-20 minute long video.

There's also just "look at this thing" videos like "$1 to $10,000,00 car" where him and his friends check out increasingly expensive cars until they eventually get a whole bridge cordoned off to drive in the most expensive car in the world.

He does some philanthropy, like his "plant 10 million trees" campaign and makes money through sponsorship deals and advertising his own brands - they're currently running their own line of (fair trade?) chocolate bars that are available (in most places?) in the US, which kids will buy because of the brand recognition, leaving them with a ton of profits.

load more comments (4 replies)
[-] [email protected] 26 points 7 months ago

I believe he is a Youtuber. That's as far as I've gotten.

[-] PocketRocket 9 points 7 months ago

If memory serves (being knowledge I gleaned from a podcast). He's a YouTuber that has carved out a popular niche in philanthropy of sorts. All for views of course, but some philanthropy none the less. Very popular I think with, I want to say Gen Alpha aged kids. A lot of people have imitated the content style in the last few years. So I guess there is instant brand recognition and trust there for a lot of people.

load more comments (3 replies)
[-] Matriks404 38 points 7 months ago* (last edited 7 months ago)

The more I hear about AI-generated content and other crap that is posted online these days, I wonder if I should just start reading books instead, maybe even learn to play on a musical instrument and leave virtual world altogether.

[-] snek 10 points 7 months ago

Plagiarized books entirely written by an AI are nor far-fetched. Get ready for a shitty reality.

load more comments (2 replies)
[-] [email protected] 6 points 7 months ago

There is already books written entirely by AI too, AI is litterly getting everywhere

load more comments (5 replies)
load more comments (1 replies)
[-] [email protected] 30 points 7 months ago

Butlerian jihad sounds like a good idea rn

[-] camr_on 14 points 7 months ago

Yeah but then I need to hire a mentat to keep my shit straight

[-] [email protected] 12 points 7 months ago

No.

We need the AIs to make the Men of Gold so we can compete with the murder orgy space elves.

[-] rez_doggie 29 points 7 months ago* (last edited 7 months ago)

Came to watch a fake mrbeast left dissatisfied. Came back to post a link: https://twitter.com/MrBeast/status/1709046466629554577

[-] [email protected] 12 points 7 months ago

Click blewlowaugh now

[-] Asudox 14 points 7 months ago

And that is why we need a pixel poisoner but for videos.

[-] [email protected] 19 points 7 months ago

I’m not familiar with the term, and Google shows nothing that makes sense in context. Can you explain the concept?

[-] [email protected] 10 points 7 months ago* (last edited 7 months ago)

Here specifically it's a technique to alter images that makes them distorted for the "perception" by generative neural networks and unusable as training data but still recognizable to a human.

The general term is https://en.wikipedia.org/wiki/Adversarial_machine_learning#Data_poisoning

One example of a tool that does this is https://glaze.cs.uchicago.edu/ but I have doubts about its imperceptibility

[-] [email protected] 9 points 7 months ago

Yeah I'm at a loss aswell. Is it a way to prove the source of a video?

load more comments (7 replies)
[-] [email protected] 12 points 7 months ago

This is the best summary I could come up with:


TikTok ran an advertisement featuring an AI-generated deepfake version of MrBeast claiming to give out iPhone 15s for $2 as part of a 10,000 phone giveaway.

The sponsored video, which Insider viewed on the app on Monday, looked official as it included MrBeast's logo and a blue check mark next to his name.

Two days ago, Tom Hanks posted a warning to fans about a promotional video hawking a dental plan that featured an unapproved AI version of himself.

"Realism, efficiency, and accessibility or democratization means that now this is essentially in the hands of everyday people," Henry Ajder, an academic researcher and expert in generative AI and deepfakes, told Insider.

Not all AI-generated ad content featuring celebrities is inherently bad, as a recent campaign coordinated between Lionel Messi and Lay's demonstrates.

"If someone releases an AI-generated advert without disclosure, even if it's perfectly benign, I still think that should be labeled and should be positioned to an audience in the way that they can understand," Ajder said.


The original article contains 518 words, the summary contains 168 words. Saved 68%. I'm a bot and I'm open source!

[-] [email protected] 11 points 7 months ago

Can Mr Beast sue til tok over this?

[-] JustZ 6 points 7 months ago

No, can only DMCA takedowns.

If he can find the creator or publisher, he can sue for misappropriation of name and likeness. It's a privacy tort.

load more comments (1 replies)
load more comments (2 replies)
[-] [email protected] 11 points 7 months ago

I read this title last night and thought it was a story about AI making an amalgamation of MrBeast and Stephen Hawking to shill iPhones

load more comments (1 replies)
[-] foggy 8 points 7 months ago

It would be a good time to start a paid online community that has a lengthy vetting process for accuracy and authenticity of all content.

load more comments (10 replies)
[-] Moc 7 points 7 months ago

Link to video?

load more comments
view more: next ›
this post was submitted on 04 Oct 2023
516 points (96.9% liked)

Technology

54592 readers
5657 users here now

This is a most excellent place for technology news and articles.


Our Rules


  1. Follow the lemmy.world rules.
  2. Only tech related content.
  3. Be excellent to each another!
  4. Mod approved content bots can post up to 10 articles per day.
  5. Threads asking for personal tech support may be deleted.
  6. Politics threads may be removed.
  7. No memes allowed as posts, OK to post as comments.
  8. Only approved bots from the list below, to ask if your bot can be added please contact us.
  9. Check for duplicates before posting, duplicates may be removed

Approved Bots


founded 11 months ago
MODERATORS