“AI is a tool, not a creature.” (Sam Altman)
AI – Artificial Intelligence – has been making waves in photography in recent years, revolutionising the way we approach the art of capturing images. But how does a modern photographer approach this area and what is the way forward for exif data and competitions?
One of the most significant ways in which AI is being used in photography is through the automation of image processing. With AI-powered algorithms, photographers can now automatically adjust exposure, contrast, and color balance to create more visually appealing images. This helps amateur photographers take professional-looking images, without having to worry about complicated post-processing techniques. Additionally, AI can help professionals speed up their workflow by automating repetitive tasks, allowing them to focus on more creative aspects of their work.
Another way in which AI is transforming photography is through the use of machine learning algorithms. By analysing massive datasets of images, AI can learn to identify and differentiate between different objects and subjects. This can be used to create highly customized image recognition tools that help photographers identify certain scenes or objects and make more informed decisions about how to capture them.
AI is also being used to enhance the quality and accuracy of photographs. With deep learning algorithms, AI tools can automatically identify and remove unwanted artifacts and noise that can detract from the quality of images. Additionally, AI can be used to enhance the level of detail and sharpness in images, creating clearer and more vivid photographs. Therefore AI is quickly becoming an essential tool for photographers of all levels. Whether you are an amateur or a professional, AI-powered image processing and recognition tools can help you take better photos, faster, and with greater accuracy and detail. As AI continues to evolve, we can expect to see even more exciting developments in this field, leading to a new era of photography that is more creative, efficient, and impactful than ever before.
AI generated image by Susan Cipriano from Pixabay.
Earlier this year, arts writer and assistant professor of visual culture at Montclair State University, Charlotte Kent, wrote an article entitiled How Will AI Transform Photography? In this she explained that:
“Artists were initially intrigued by an early form of AI known as GANs, or generative adversarial networks, which created composite visuals from existing data, including images. GANs allow artists to compile personalized data sets; for instance, photographers may use their own archive of images to create novel work. In contrast, text-to-image generators depend on massive data sets that no single artist can compile. Until recently, they also needed invitations from developers to access these generators. But, by 2022, with the proliferation of the technology, a growing number of artists began sharing their experiments, and people saw first hand what these systems could create from a simple string of words.”
Should AI photo’s be differentiated from human assisted or Raw photos?
Photographs made with AI are becoming increasingly popular due to their ability to enhance images and produce high-quality results. In contrast, raw photographs are unprocessed, capturing the image exactly as it appears at the time of capture. While both types of photography have their advantages, there is a growing debate over whether photos made with AI should be differentiated from raw or human assisted photos. Some argue that AI-assisted images should be clearly labeled as such, while others believe that the distinction is unnecessary.One argument in favor of differentiating AI-assisted photographs is that they can give a false impression of reality. AI algorithms can manipulate images to change colors, add or remove objects, or even create entirely new images. While this can be useful for artistic purposes, it can also be misleading if someone believes they are looking at an unaltered photograph. However, others argue that the purpose of photography is to capture a moment or tell a story, and the use of AI can enhance this.
AI or Artificial Intelligence is not the invention of a single person, but rather a field of computer science that has evolved over several decades, with contributions from numerous researchers, scientists, and developers. In the 1950s, pioneers such as John McCarthy, Marvin Minsky, and Claude Shannon made significant contributions to the development of AI. Since then, the field has continued to grow and evolve, with significant contributions from researchers all over the world.
By enhancing colors, adjusting focus, or improving sharpness, AI-assisted photographs can provide a clearer representation of the moment captured. Ultimately, whether or not AI-assisted photos should be differentiated from raw photos is up for debate. While there are valid arguments on both sides, the most important thing is for photographers to be transparent about their techniques and for viewers to be aware of the potential limitations of AI-assisted images.
However, the problem that is now raising its head is that AI is being used in many photography software packages.
Professional Photo magazine had a brief look at some of these packages in an article in issue 203, that included Imagen AI which has paved the way to an entirely new way of working, by providing a full edit based on its AI engine learning how you’ve previously edited your photos, creating an individual profile to suit your signature style, and then batch applying that to any work you process through its system; PortraitPro, which has just released its latest version, has changed the editing game for this type of work; Nvidia Canvas which uses AI to turn simple brush strokes into frighteningly realistic landscape photographs (left), and Midjourney which is so new it’s still in its Beta stage, and it’s only accessible through Discord, which is a server-based community. With Midjourney you can create any unique texture or image you want, and the best thing is that it’s so fast, and it provides incredibly detailed results. We input a prompt of ‘landscape, autumn, leaves, realistic,’ and it gave the results you see to the left. You can even upscale them to get higher resolution versions that will then be usable in photo art.
Then there is Adobe Photoshop’s AI-powered tools, such as Adobe Sensei, which can be used for tasks like selecting and masking objects, automatically enhancing images, and generating alternative compositions. Skylum Luminar uses AI-powered editing tools to enhance images, such as its AI Sky Enhancer and AI Structure tools. Topaz Labs offers several AI-powered plugins for photo editing software, such as Topaz AI Gigapixel for enlarging images without losing quality and Topaz Sharpen AI for correcting blurry images. Google Photos uses AI to automatically organize and enhance photos, such as by adjusting brightness, contrast, and color balance. DxO PhotoLab has AI-powered features such as DxO Smart Lighting, which can automatically adjust exposure and contrast in images. These are just a few examples of photo software that utilize AI to improve the editing and enhancement process. As AI continues to develop and improve, it is likely that more photography software will incorporate AI-powered tools and features.
There are several AI image creation software packages which use machine learning algorithms to generate realistic images of people, landscapes, and objects. Some of the popular ones include DeepArt.io, NeuralStyler, and Prisma. These tools use deep neural networks to analyze and transform input images into various styles and formats. They offer features such as image resizing, color adjustment, and texture mapping, and provide a user-friendly interface for easy customization. The choice of software depends on the specific requirements of the project, the quality of the output images, and the cost of the tool.
So how would we define a photograph in these cases, where the human is assisted? Should there be a grading system to indicate how much AI a particular photo may have used? Would this need to be encoded into the exif data to ensure transparency and accuracy in the representation of the photograph? Some proposed AI metadata tags could include information about the type of AI used, the purpose of the AI, and the degree of AI involvement in the image creation process. However, as of now, these tags have not been widely adopted or standardised. If implemented metadata should be added in a clear and understandable way, and there should be clear guidelines and standards for what kind of information should be included. This information could also be useful for photographers, who could use the embedded metadata to keep track of the AI techniques they used in creating their images and to help them improve their photography skills.
AI in Photography Competitions
It is important for photography competitions to establish clear guidelines and rules regarding the use of AI in photography. For example, competitions could require that photographers disclose the use of AI in their entries, and that any images entered into the competition have not been heavily manipulated or created entirely through AI algorithms.Judges should also be educated on the use of AI in photography and be able to recognize whether an image is mostly original or mostly AI-generated. Additionally, photo competitions should consider creating separate categories for AI-assisted photography, as it may not be appropriate to compare these images directly with raw photographs. This could help ensure that all photographs are judged on a level playing field and that the use of AI is not unfairly advantaged or penalized. Overall, it’s a possibility that AI photos should be allowed in photo competitions as long as they are evaluated fairly and transparently and that judges and organizers establish clear guidelines for their use.
Obviously, there are also concerns about the potential for AI to replace human photographers altogether. While AI can be used to automate certain tasks and increase efficiency, some worry that it could lead to the devaluation of the human eye and creative vision in photography, and that would hit the bottom line, the livelihood of photographers themselves. Whichever method is used, one advantage of human photography over AI is that humans have creativity and can bring a unique artistic perspective to their photography. Humans can also use their intuition to capture moments that may not be detectable by AI. Additionally, humans can adjust their photography techniques and equipment to adapt to specific situations, such as low light or fast-moving subjects, which AI may struggle with. Finally, humans can establish personal connections with their subjects, which can lead to more meaningful and emotionally powerful photographs.
The Exchangeable Image File Format (EXIF) guidelines for digital photography were created by the Japan Electronics and Information Technology Industries Association (JEITA) in the early 1990s. The EXIF standard was developed to provide a standard format for storing metadata information in digital photographs. Since then, the standard has been updated several times and is now maintained by the Committee for the Standardization of Information Technology of the Japan Electronics and Information Technology Industries Association (JEITA/CSIT). While JEITA/CSIT is responsible for maintaining the EXIF standard, many software companies and camera manufacturers also contribute to its development and implementation.
As photographer Sean Tucker explains in the video at the end of this article, CGI has already replaced photographers in some cases, such as where Ikea use it for their kitchen images. But in places where it is essential that real photographs are used, such as in news reporting or street photography, AI would naturally have to be be banned. Anything really, where a true story is being told
.
AI image generated from Sam Altman’s words AI “is a tool, not a creature”, by Dane Pestano
EU, US and British Regulation of AI
On the 16th of May 2023 a Senate committee hearing took place in the USA to discuss the broader implications of AI and its potential threats. Being questioned was the chief executive of OpenAI, Sam Altman, whose company created the app ‘ChatGPT’ used in writing some of this article. Altman says there should be a new copyright model for art to compensate artists whose art was used to create something new. “Creators deserve control,” he said. Altman also said he believed regulation should demand images state when they have been generated by AI.
Sam Altman
Some key moments noted by the BBC:
- Richard Blumenthal, chairman of the Senate subcommittee on privacy, played a statement manipulated to sound like him by AI as he opened on the potential harms of the technology.
- OpenAI CEO Sam Altman said that AI has the potential to solve “humanity’s biggest challenges, like climate change and curing cancer”.
- Altman and his fellow witnesses agreed that government regulation is “crucial” to the future of the industry.
- AI-skeptic professor Gary Marcus said the US and others had “acted too slowly with social media” regulation but has choices to make right now regarding AI.
- Marcus proposed a new cabinet-level agency to regulate the sector, an idea that Altman also appeared to back.
- Altman said AI advancements will impact the workforce significantly but current models remain “a tool, not a creature”.
- Asked about AI’s impact on music, Altman said that content creators should have a say in how their voices, likenesses or copyrighted content are used to train AI models.
- Christina Montgomery, IBM’s chief privacy and trust officer, said the EU’s rules on AI are “regulating by context” and provide a good lead for the US to follow.
- Blumenthal said AI companies should lead with a “do no harm” approach but conceded there would be no pause in AI development until regulators can catch up. (End BBC).
The proposed EU regulations on AI aim to regulate AI systems by context, meaning that the level of regulation required for an AI system will depend on the potential risks associated with its use. For example, high-risk AI systems, such as those used in healthcare or transport, will be subject to stricter regulations than low-risk AI systems. The proposed regulations also take into account the intended use of the AI system, as well as the potential impact it could have on individuals and society as a whole. By regulating AI by context, the EU hopes to ensure that AI systems are developed and used in a responsible and trustworthy manner, while also promoting innovation and growth in the field of AI.
As to Britain, the British government has established ethical AI guidelines under the purview of the Centre for Data Ethics and Innovation (CDEI). The guidelines aim to ensure that AI is developed and deployed in a way that is transparent, accountable, and fair. Additionally, the UK has laws such as the General Data Protection Regulation (GDPR) and the Data Protection Act 2018 that regulate the collection, use, and storage of personal data by AI systems. Overall, the British government is taking steps to ensure that AI is developed and regulated in a responsible manner.
As Alex Hearn and Dan Milmo have noted in their article on the 7th May 2023 in The Observer – All you need to know about AI – the biggest and most overt call for restraint is coming from AI Professionals. They comment that: “in March Elon Musk, a co-founder of OpenAI, was among the signatories to letter calling for a pause in major AI projects.” A global framework is required, but how long that will take is anyone's guess.
Tested – Retouch4Me
ACDSee launches Photo Studio 2025
Is AI the Devil or a Godsend?
British Telecom have recently announced the loss of 55,000 jobs over the coming years as it implements AI technology in its customer services call centers. Evidence of a worrying trend of job losses.Professional Photo Magazine has recently run articles on the advent of AI hardware and software, with the cover photo of issue 203 being AI created by Giles Christopher, so it has arrived in the mainstream. In the main article we mentioned that in terms of AI technology, we’re still relatively early on in the process and, given how realistic results already look, the fear will be that the programs out there will continue to be refined and they’ll be joined by other, still more powerful, applications. So, the foreboding experienced by some is perhaps justified, and Giles has experienced first-hand unfavourable responses and resistance to the work he’s outputting:
Also, as Charlotte Kent comments, “These are still early days, and many feel this is a special, even fragile time for AI-generated imagery. To discredit these representations based on image quality or veracity alone is to miss the larger point about the inequality of the culture from which they emerge—both in terms of social mores and online data. Experimenting with these systems is a first step toward demanding they be made better and more inclusive.
[…] AI may have already begun radically altering how we think about photography, but it’s helpful to remember that such discussions have always been a part of the medium’s evolution. Older arguments around authorship and creativity willreappear in this new context, and artists who consider the aesthetics and ethics of images are right to demand more conscientiously collected data. As AI’s relationship to photography continues to be defined, it will nonetheless be shaped by the artists exploring the technology’s limits in playful, inventive, and critical ways. “
As Altman says, “AI is a tool, not a creature”, so creativity still all starts with our prompts, but for how long?
95% of images in this article and 90% of the text were created using AI.
