You have robot vision if you can see what's strange about these videos within 30 seconds
PEOPLE have been left 'terrified' by a series of seemingly ordinary videos.
But can you tell what's wrong with it within 30 seconds?
2
“I watch every video on that website in disbelief,” one X user wrote after discovering the truth behind it.
“This is wild,” another person wrote.
The truth is that none of these clips are real: they are all created by AI.
OpenAI, the company that shocked the world with ChatGPT, has released a more advanced version that can now generate videos.
Like previous image-generating models, a person just needs to describe what they want in writing and the AI will create it immediately.
The technology – known as Sora, the Japanese word for 'air' – can produce realistic images up to a minute long.
The idea that an AI can create a hyper-realistic video of, say, a politician doing something untoward should set alarm bells ringing as we enter the most election year in human history.
Dr. Andrew Rogoyski
It is also able to create a video from an image or expand existing footage with new material.
But just like the AI-generated images we've seen with extra fingers and teeth in recent years, some videos need some work.
Take the example below from OpenAI of a cat waking its owner and exhibiting two strange errors. Can you discover them?
First, the woman's arm disappears when she turns around.
And second, the cat gets an extra arm and paw on the left side.
Hazards
The service is not yet available to the general public.
For now, only selected researchers and creators are invited to test it for dangerous flaws such as misinformation, hateful content and bias.
Dr. Andrew Rogoyski from the University of Surrey said: “OpenAI has recognized the potential harm of such a system.
“The idea that an AI can create a hyper-realistic video of, say, a politician doing something untoward should ring alarm bells as we enter the most election-heavy year in human history, with more than 60 democratic elections in '24 and a half.” the population of the planet votes.
“Interestingly, OpenAI plans to watermark Sora's output with C2PA, a digital certification system that is increasingly popular as a means of tracking the provenance of information.”
In a web post, OpenAI explains: “We will engage policymakers, educators and artists around the world to understand their concerns and identify positive use cases for this new technology.
“Despite extensive research and testing, we cannot predict all the beneficial ways people will use our technology, nor all the ways people will misuse it.”
Read more about Artificial Intelligence
Everything you need to know about the latest developments in artificial intelligence