Artificial Intelligence is undeniably here to stay. While the technology offers immense benefits, it also comes with significant risks and challenges. In essence, when used responsibly, AI has the potential to achieve remarkable outcomes, but in the wrong hands, it could lead to catastrophic consequences. Advances in AI technology in medicine and engineering has proven to be successful in those fields specially improving the lifestyle of people with disabilities. As a pilot I embraced the introduction of AI into live traffic in Microsoft Flight Simulator or the AI copilot to handle communication and setup navigation when flying complex airplanes.
When it comes to my photography and videography, I have seen the benefits of this technology in many ways. Adobe’s Camera Raw with built in AI technology accelerate the retouching process by creating better subject selections and removing unwanted objects in the most natural way, features that are very convenient when shooting commercially. On my wildlife photography I can generate softer backgrounds reminiscent of large aperture lenses and can even selectively chose the focus areas after the image was taken. These are feature that really blow your mind and make your time behind the computer easier and shorter. From the healing and cloning capabilities to the AI detection and masking of individual people in a shot, Adobe has improved the way we do photo retouching.
On the flip side, the ability to generate images from text commands has significant drawbacks. Firstly, it overlooks the importance of camera features and the expertise of the photographer. Secondly, it often produces images with glaring flaws, such as people with extra fingers, birds missing legs, or elephants with zebra stripes. The disturbing aspect is that many people come across these AI-generated images of imaginary birds on platforms like Facebook and believe them to be real. This illustrates the darker side of AI in photography, where virtually anything can be recreated by AI and passed off as genuine to the unsuspecting viewer. Even worse, when people’s images are manipulated with this technology to spread misinformation on various topics.
Some companies, including Adobe, have faced criticism for misleading their customers about how they obtain images to train their AI engines. This can be quite upsetting for those who pay a monthly subscription for Adobe Photoshop or Lightroom and unknowingly agree to the consent during installation. Ultimately, AI’s effectiveness hinges on the quality of the information it receives—be it images, text, or voice. The outcomes will always depend on the intent and usage.