With technology continuing to advance at breakneck speed, there has been a lot of talk lately about whether digital rendering can really replace product photography. Taking it a step further, could artificial intelligence one day simply generate images without any input from the photographer or digital artist?
As photographers, we’re often amazed at how amazing modern technology can be, how magical new “must have” camera features or how intelligent image processing software has become. how. I don’t consider myself particularly old, but thinking back on using a manual focus 35mm SLR (because that’s all we had to use, not because I’m a connoisseur) and compared Compare that experience with the incredible face detection or eye detection autofocus on modern mirrorless cameras, it’s hard to believe that these huge technological advancements have happened in a lifetime. I. Even the act of sitting in my living room, controlling the lighting and home entertainment with my voice, or video calling a friend in another country on one iPad are actually things that my kids themselves would consider science fiction. Even my smartphone is significantly more powerful than any computer I used before growing up. In the grand scheme of human history, it’s time for us to move from the first commercial camera available to the general public to a smartphone with a very efficient digital camera in most people’s pocket. in the western world, it’s amazing.
DALL · E mini
This week, I saw some hilarious images posted on social media from a project called DALL · E mini. Those are crude images of very random things like Joe Biden eating hot dogs or spiders in felt coats. Silly images on the internet are nothing new, but they are believed to have been created by artificial intelligence. Some of these images look like simple drawings or cartoons, while others look like raised renderings from a video game in the early 2000s. None what I consider to be fact. Most of the images shared on social media are pretty silly, so naturally, I wanted to look up where these images came from. A short Google search brought me to DALL · E mini. This Transformer-based text-to-image model was designed by Boris Dayma, Suraj Patil, Pedro Cuenca, Khalid Saifullah, Tanishq Abraham, Phuc Le, Luke, Luke Melas and Ritobrata Ghosh.
DALL · E mini very simple. You enter a short text prompt, then the AI, which has been trained on unfiltered data from the internet, goes to work and generates nine images based on the text stimulus. These images often vary widely, but represent the AI’s interpretation of your input, based on data on the internet. Right now, it’s not particularly fast, taking anywhere from two to four minutes to produce the best quality images. After playing for longer than it should have, I can see it being nothing more than a meme goldmine right now, but conceptually it’s very engaging, with exciting future possibilities.
This model is intended to be used to generate images based on text prompts for research and personal consumption. Intended uses include supporting creativity, creating humorous content, and providing generations of curious people about the model’s behavior. The intended use excludes those described in Misuse and out of scope section.
It is worth noting that these images are generated by artificial intelligence, trained on unfiltered data found on the internet, to generate unique interpretations for the search terms that users search for. presented as a selection of basic images. It’s also worth considering that people on the internet are using their own creativity and imagination to ask this AI to create things to humorous effect. If you’re going to see the discussion board or try out the image generator for yourself, be aware of the biased and restrictive text provided by the developer team, and be aware that some people on the internet are headliners. It’s funny when the AI creates problematic or offensive images.
While the ability of models to create images is impressive, they can also reinforce or exacerbate social biases. Although the extent and nature of the biases of the DALL·E mini model have not yet been fully documented, due to the fact that the model was trained on unfiltered data from the Internet, it has may create images containing prejudices against minorities. Work is underway to analyze the nature and extent of these limitations.
The future of AI-generated images
It’s probably safe to say that no photographer will ever lose his job to AI anytime soon. However, the technology raises some questions about what the future of images might look like. Now, we live in a world where stock of images is available online in seconds to anyone who needs a generic image. Sure, the archival images are taken by a creative expert who will earn some income from them, but what happens when the machine learns to the point where some generic images can be generated by WHO? Who owns the rights to those images? Could this one day displace a large part of the archival imaging industry and be harmful to stock gallery photographers? Could we one day see renderings of products or places produced entirely by machine algorithms used for commercial purposes?
Silly memories and images aside, I wanted to see how close this system was to creating a lifelike landscape, so I included DALL · E mini a simple text input to see it would look as beautiful as a text input. This is the image it created this morning. Beware, landscape photographers! Machines are coming for your job!
I appreciate that from the look of the images created today, it seems a bit overwhelming to think that it can replace a professional photographer, but 30 years ago, iPads and FaceTime were the stuff of science. science fiction, but now we all carry a tiny computer power with a high megapixel digital camera in our pocket every day. Future possibilities are exciting or scary, depending on your point of view.
Rendering, still created by humans, is replacing the word product photography in some places. Is it only a matter of time before digital images are so lifelike that we won’t need actual photographers in many situations? Is it possible that one day there will be no need for advertising photography?
What do you think about rendering or AI replacing photography? Is this technology interesting or disturbing? How far can legitimate commercial use for this technology go? Let me know in the comments section.
