HomeTech The images of the floods in Spain were not created by AI. The problem is that people think they were

The images of the floods in Spain were not created by AI. The problem is that people think they were

0 comments
The images of the floods in Spain were not created by AI. The problem is that people think they were

METROA striking photograph caught my eye in the most recent issue of Charles Arthur’s Substack newsletter. Social warming. It shows a narrow street after the “rain bomb” that devastated the Valencia region of Spain. A year’s worth of rain fell in a single day and in some locations more than 490 liters per square meter fell in eight hours. Water is very heavy, so if there is a slope it will flow downhill with the kind of force that can lift a heavy SUV and throw it like a toy. And if it flows down a narrow urban street, it will throw up parked cars like King Kong in a bad mood.

The photograph in Arthur’s article showed what had happened on a specific street. Taken with a telephoto lens from the top floor of a building, it showed a chaotic and almost surreal scene: about 70 vehicles of all sizes in disarray and scattered at crazy angles along the street.

It was an amazing image that really stopped me in my tracks. Not in vain, it also went viral on social networks. And then came the reaction: “AI image, fake news.” The photograph was so vivid, so incredibly sharp and unreal, that it looked to viewers like something they could have faked using Midjourney or Dall-E or a host of other generative AI tools.

But it wasn’t fake, as Arthur demonstrated in some nice detective work. tracing a bar in the image use Facebook, find the street in Apple Maps and even “walk” along it using Street View. “It’s unclear why these people thought that particular photo wasn’t real,” he writes. “Maybe it’s something to do with the shine of the cars and the peculiar roundness of the shapes, and maybe the lack of obvious damage.” Or is the proliferation of AI-generated fakes already making people increasingly predisposed to not believe in real things?

My hunch is that it’s the latter, because social media is being invaded by what has become known as “AI creep”: images and text created using generative AI tools. (Amazon’s Kindle store is have similar problems with AI-generated “books,” but that’s a different story.)

One would have thought that social media companies would be bothered by this tsunami of trash on their platforms. Think again. According Jason Koebler of tech news website 404 MediaIn a recent quarterly earnings call that was overwhelmingly about AI, Meta CEO Mark Zuckerberg said new AI-generated feeds were likely coming to Facebook and other Meta platforms. Zuckerberg said he was excited about the “opportunity for AI to help people create content that simply improves people’s experience.”

Enthusing about his topic, Zuck continued: “I think we’re going to add a whole new category of content, which is content generated or summarized by AI or a type of existing content put together by AI in some way. And I think that’s going to be very exciting for Facebook and Instagram and maybe Threads or other types of feed experiences over time.”

Which, in a way, makes a lot of sense: Meta’s profits depend on keeping users of its platforms “engaged” (i.e., spending as much time as possible) and if AI helps achieve that goal, what is the problem?

On the supply side, it turns out that AI-generated material is also profitable for those who create it. Koebler has spent a year exploring this dark side of social media. In India, he met Gyan Abhishek, an analyst who studies online virality. Abhishek showed him a striking image that was being used to generate income: the image of a skeletal old man hunched over while being devoured by hundreds of insects.

“The Indian public is very excited” Abhishek explained.. “After seeing photos like this, they ‘like’, ‘comment’ and share them. So you too should create a page like this, upload photos and earn money with the performance bonus.” He also claims that creators of viral images can earn $100 per 1,000 likes, which sounds like improvisation money, at least to this columnist.

So what we have here is a nice positive feedback loop in which AI creators benefit by feeding the engagement algorithms of social media platforms, which in turn benefit from the increasing “engagement” that viral images attract. . The problem with positive feedback loops, however, is that they lead to runaway growth and the question of what happens to social networks when they become terminally dirty as a result. Where Meta and company are going.

skip past newsletter promotion

what i have been reading

How fragile is autocracy?
A sobering assessment of Trump’s and America’s victory… What the future looks like from here – by Professor Dave Karpf of George Washington University.

Click here for more
An insightful essay by Jason Kottke about the art of writing for the web and the power of hyperlinks to intensify a discussion.

The musk effect
Machiavelli of the market – a timely essay on the enormous power of entrepreneurs written by Alex Gourevitch.

You may also like