Forget deepfakes, the new photo editing AI is beyond imagination

Technology continues to evolve. AI-generated content has now become deceptively real. At first glance, many of the custom-made Photos indistinguishable from actual images. So be careful what you find on the web. You should not blindly trust your eyes with every job. The creations of this new photo editing-AI are vastly beyond imagination.

Other AI in action DALL-E 2

DALL-E 2 is another AI, which has gained attention for generating images on the Internet. Similar to and Artflow.ai or google image generates content based on users’ text input. This creates realistic images and works of art in which the AI ​​has been programmed using learning content. DALL-E 2 is a continuation of DALL-E and works as a project to investigate machine learning. Therefore, no commercial use of this software is anticipated in the future.

However, it is possible to use them for private purposes, and this could lead to more problems. DALL-E 2 limits the access of new users to a certain test quota. Ultimately, however, it cannot be ruled out that one of the future users will use the generated content somewhere for questionable purposes. The project itself seems to be aware of this, since it reserves the right to reject requests from users that present such a risk of use. This can at least reduce the risk, if not prevent it.

New photo editing AI with great power

The performance of DALL-E 2 is impressive. The new image processing AI can generate realistic 8K shots of animals or places that are difficult to distinguish from an original. The creation of many generated artworks, which are based on the drawing style of other paintings, is also easily possible. To achieve this, the operators of the DALL-E 2 software have provided a large amount of training material to the Alabama teach to recreate important characteristics of people, animals, places and works of art.

The type of connection with the words is also decisive. Finally, the AI ​​must understand what users expect of it whenPhotos‘, ‘dogs’ or ‘paintings’. The potential offered by DALL-E 2 is very promising. But precisely because of the realistically generated content, the risk of someone misusing the created images could be great. The DALL-E 2 itself provides each generated image with a signature color code in the lower right corner. However, it is not difficult to remove them and reduce the image if someone wanted to use the product for other purposes.

Not the fault of the AI ​​for its intended use

Despite all the criticism of AI content generation or modification, it should be clear that the technologies themselves do not determine their intended use. With the help of machine learning and AI, many useful and useful creations can be created. Unfortunately, however, there will always be people who use such a tool in an attempt to gain an advantage. Whether someone else’s doctored recordings appear as a deep fake to harm the person or a doctored fundraising campaign collects the money from many bona fide helpers.

That is why it is even more important that you remain vigilant and skeptical about the content on the Internet. How likely is it that the represented person actually does what is represented? Is there proof of a promoter or owner of the page where the content is available? Be suspicious if your instinct gives you suspicions or if you come across something unexpected. Pay close attention to proper lighting conditions in images and look for discrepancies. As creatures that process a lot of visual data, we often already have a subconscious idea that something is wrong. However, what exactly is irritating with a rendering often only becomes clear on closer inspection. You should also be skeptical about video recordings if only very little vocabulary is used and stiff facial expressions are presented.