Last year, Metro Night News editor Bernie Davis wrote a story about the piano that appears on London's railway station platforms.
The commuters were confused, to say the least, but apparently they didn't care about the music intertwining.
Except Davis didn't – he didn't subway In 2024, commuters at Clapham Junction were not “confusing” by “mystery piano” given that this never happened. (Davis can also spell words correctly.)
This is a fake news story that I asked ChatGpt, an artificial intelligence (AI) chatbot, to generate today. It only took a minute or two to make it.
Over 34 million AI images are generated over a day, and technology is constantly improving the production of realistic photos and videos.
“AI images not only fake reality, they not only bending context, but can drive dangerous false narratives even innocuously-looking photos,” said Naomi Owusu, CEO and co-founder of the live digital publishing platform. subway.
“When fake images are seen, you can install a 'new reality' in people's minds. Even if it proves false later, it can be very difficult to undo the impression it leaves behind. ”
The expert said subway For now, at least there are still quite a few ways to find fake AI.
1. It's all in detail
Images created using text-to-image algorithms look good from afar, but often far from Vaidotasšedys, the Chief Risk Officer of Web Intelligence Solutions providers. Oxilabsaid.
“Hands are often a prize,” said šedys. “Too many false or strange shadows on fingers, limbs suggest that the image is not human.”
I'll take the fake subway In the story, the copy has a spelling error (“Flex size” and the favourite section “e-dutton”) and the font is filled.
Or, the tickk clips of the virus above show bunnies bouncing off the trampoline, so the furry animals blend in with each other and sometimes distorted.
Julius -Cerniauskas, CEO of Oxylabs, added:
2. “Robot” lighting
Well, first of all, there is no judgment here. All of us I've been fooled By AI images. It's not just youBut his parents, politicians, CEOs. result? That means that the technology is doing exactly what it was designed to do: Looks authentic.
The above is an attempt to write chatbots as they are now infamous, in the style in which šedys is “overly refined robots.”
“Look for content that lacks a natural rhythm or relies on too many common examples,” he adds.
3. Deepfake vs AI
Deepfake Technology is software that allows people to exchange faces, voices and other characteristics to create digital dolls to speak.
“Some people are developing deepfakes for fraud and extortion, and producing 'proof' that compromised to deceive others,” Cerniaskas said.
'In other cases, deepfakes are often used to spread misinformation with political motivation. They can destabilize the country, shake public opinion and be manipulated by foreign agents.
“Even people who create deepfakes just for the thrills don't take into account the damage they may cause.
“Just because we're not a victim of deepfakes doesn't mean that it won't happen because AI continues to improve.”
4. Use reverse image search and other tools
ChatGpt and other synthetic image makers are great at Faces as examples of generative AI, so drink online data.
Researchers found that faces created by AI systems are perceived as more realistic than photographs of people called hyperrealism.
Ironically, a prize where the image is not real is that someone's face looks a little too perfect.
Owusu said: “No longer trusting your eyes, it's about checking the source. You can't fact-check just pixels.
“Use common sense: ask yourself how possible it is that what you are seeing is real, and how likely is it that someone has captured that exact moment. If Photoshop is too visible, it could be AI.”
If in doubt, do a reverse image search to find other locations on the internet where the photo resides.
“If it cannot be tracked, it should be treated as untested, regardless of how convincing it is,” Owusu said.
Services like Zerogpt can use rifles via text to see if it's an AI Jibber Jabber.
Check out the metadata that is a digital fingerprint embedded in photos, documents, and web pages.
5. Curate your social media feed
According to an analysis by Garbage Day Newsletter, in May, four of the top 10 YouTube channels by subscribers featured AI-made material in all videos.
Some AI content creators don't specify that what their followers are watching is audio, whether they say it in captions or click on the “made with AI” tag on social media platforms. Some accounts are merely AI bots and will spit out fraudulent posts and images.
This shows how important it is that all three experts we said that all three experts are following, and that it is especially important that they are intentional to the news outlet.
Why is this important?
Free and easy to use, countless people now write emails, plan your weekly budget, or Get life advice.
But many people also create indemnity from flooding fake photos and videos about world leaders, political candidates and their supporters. This misinformation conveyor belt could shake elections and erode democracy, political activists said earlier. subway.
The International Panel on Information Environment, a group of scientists, found that AI was used in 80% of more than 50 elections last year.
It won't take long until it becomes “impossible” to distinguish the fake images from the video from the real images, Cernia Wuska said.
“You shouldn't take it for granted that you can find a difference from the real thing,” he added.
Governments and regulators may try to grasp AI hoaxes, but constantly blessed technology can easily outweigh the law.
“Don't trust everything that appears on the screen,” šedys said. “As AI continues to evolve, critical thinking will become your best defense.”
Please contact our news team by emailing webnews@metro.co.uk.
For more stories like this, Check the news page.
Details: Parents and children use ChatGpt for their studies – Are AI raising a generation of “tech-dependent empty heads”?
Details: The reason why people say “please” and “thank you” is 158,000,000 bottles of water per day
More Information: Everything you need to know about the latest chatgpt updates
