6• Do you know who—or what—created what you viewed today?
Text and images by LensAfield*
12/24/2022, amended 4/13/23
You would have to literally be living in the middle of Nowhere not have been exposed to discussions about AI (artificial intelligence). The depth and breadth of its applications and implications are already mind-boggling.
Here we discuss photography, so I will limit what I say to that topic.
So just how is AI being used in photography (and by extension, cinema and videography)?
AI is having a significant impact on the world of photography and videography, and its influence is only set to grow in the coming years. From image and video editing software to image and video analysis software, AI is changing the way photographers and videographers create and edit their work, making it easier and more efficient than ever before.
One of the key ways in which AI is impacting photography is through image and video editing software. These products use AI algorithms to help users enhance and improve their photos and videos in a variety of ways. For example, they can automatically correct color balance, remove objects from the frame, and reduce noise. This makes it easier for photographers and videographers to achieve the desired look and feel for their work, without having to spend hours manually editing each image or video.
Another area where AI is having a significant impact on photography is image and video analysis software. These products use AI algorithms to analyze and extract information from images and videos, such as identifying objects, people, and other features. This can be useful for tasks such as categorizing and organizing images and videos, as well as for identifying patterns and trends.
AI is also being used to generate new images and videos from input data. For example, AI-powered software can create realistic images or videos from 3D models, allowing photographers and videographers to create content that would otherwise be difficult or impossible to capture in the real world.
In addition to these uses, AI is also being used in image and video recognition software, which allows users to recognize and classify objects, people, and other features in images and videos. This can be useful for tasks such as automatically tagging images and videos with relevant keywords, making it easier for users to find and organize their content.
Finally, AI is being used in image and video search software, which enables users to search for specific images or videos based on keywords or other criteria. This can be especially useful for photographers and videographers who have large collections of images and videos, as it allows them to quickly find the content they need without having to manually search through their entire library.
Overall, AI is having a significant impact on the world of photography and videography, and its influence is only set to grow in the coming years. From image and video editing software to image and video analysis software, AI is changing the way photographers and videographers create and edit their work, making it easier and more efficient than ever before.
Everything you just read after the question I posed above was generated by an AI engine that I asked to produce a 500 word article on the use of AI in photography.
I then asked, “How many AI products are available for photography and videography and of what categories?” It generated:
There are many AI products available for photography and videography that fall into various categories. Some examples include:
- Image editing software: These AI-powered tools allow users to enhance, retouch, and manipulate images. Some popular examples include Adobe Photoshop, Affinity Photo, and GIMP.
- Image recognition software: These tools use AI to analyze and identify objects, people, and other elements within an image or video. Some popular examples include Google Cloud Vision, IBM Watson Visual Recognition, and Microsoft Azure Computer Vision.
- Video editing software: These AI-powered tools allow users to edit, enhance, and manipulate video footage. Some popular examples include Adobe Premiere Pro, Final Cut Pro, and Avid Media Composer.
- Video analysis software: These tools use AI to analyze and extract insights from video footage, such as identifying objects, people, and activities. Some popular examples include IBM Watson Video Analytics, Amazon Rekognition Video, and Google Cloud Video Intelligence.
- Virtual reality (VR) and augmented reality (AR) software: These AI-powered tools allow users to create and experience immersive, computer-generated environments and overlays. Some popular examples include Unity, Unreal Engine, and Blender.
Overall, there are many different AI products available for photography and videography, and the specific tools that are best suited for a particular task will depend on the needs and goals of the user.
[There will be no further AI-created text content from this point on. Do not disregard that sense you felt as you read that there is something odd about the “feel” and structure of what was created. That sense of oddness will help you to recognize AI-created text content, as will your perception of details you can’t quite describe that help you to recognize AI-generated “art” and “photos.”]
The implications are staggering. I had been thinking and reading about this topic all week and was considering writing something about it when this video link landed in my mailbox this morning and I knew it was time to do so.
There was probably some subject-related advertising prepended to the video. I hope that made you aware of just how much AI has permeated the photographic and video arts.
He’s right about the impefections in these early stages of AI development. When it comes to the creation of images from text description, the results can be disappointing. For example, I am contemplating starting a new web site titled This Old Goat about how AI is impacting photography and videography. I wanted to produce one of those little circular author images to add in the banner section or next to my bylines.
[update 4/13/23: The site This Old Goat is up and running.]
I checked out a few sites that allow the visitor to create a few test images before committing to a subscription. I asked the engine to “generate an image of an old goat eating some photographs while looking at a computer monitor.”
What I got were images of old men sitting in front of computer screens, some sprouting horns. That made me chuckle. Or simply of goats. But none of them produced a goat eating a photo while viewing a computer screen.
After about an hour of disappointing results across a few different engines, I changed my input text to ask for just “a goat in front of a computer monitor.” One engine produced this one which I kind of like,
so I edited and cropped for the use intended.
And I edited this one for the same purpose. It might better convey the drift I have in mind for the site.
Obviously, AI can generate images I can’t possibly produce of my own devices. But if you are looking to create photo-realistic results, most consumer-grade AI-generated images are generally disappointing. But there are some incredibly sophisticated engines ($$$$) out there that produce spectacular, undetectable as completely faked imagery, including faces of people that don’t actually exist, and have the ability to do some unbelievable edits and manipulations on existing photos (the implication for bad actors is enormous and more than a little frightening).
This subject is playing on my mind and is more than a little unsettling. And, apparently, is on the minds of a lot of others based on the volume of content out there on this topic. What do you think?
But before you answer, ask yourself the title question again.
*Or were they? They were only in the sense that they were created based on my text input. And by the way, that includes the banner image. And for the record, this is the first post of mine that contains images not created by me. The question is, will it be my last?