AI in Photography – Balancing Practical Benefits with Ethical and Environmental Concerns

Whether you like it or not, AI is becoming increasingly common in our everyday lives. From facial recognition on our phones to our social media feeds, in email spam filters, and navigation apps, we all use AI to help make our lives easier.

There’s no denying that AI (Artificial Intelligence) has countless positive benefits. From helping doctors to diagnose cancer to improving gender imbalance in the media, there are many ways that AI is helping improve the lives of people around the world. It’s also being used to help monitor the impacts of climate change and build better solutions.

It’s not surprising that AI tools are also filtering into our businesses too, and are quickly becoming a part of our modern photography business workflows.

However, every time you use AI, it comes at a cost – to people and the planet.

While AI can be a powerful tool, its use also raises several concerns. From the environmental impacts to the risk of bias, it’s crucial for photographers to weigh the benefits against the drawbacks. In this article, we’ll explore both the positive and negative impacts of AI tools and discuss why we should use them thoughtfully and sparingly.

Where is AI used in photography?

In recent years, AI tools have revolutionized the world of photography, and it’s seeping into almost all aspects of running a photography business, such as:

Image Processing

There is AI software for almost the entire photography editing workflow, including tools that assist with culling, editing, retouching, and image manipulation.

Copywriting

Many photographers are turning to AI tools to take over some of their most hated tasks, such as blogging, writing emails, and creating social media captions.

Marketing

From the algorithms of social media and search results to AI website design tools, AI is filtering into many aspects of marketing.

Image Generation

Many creatives are turning to AI image generators to create images for mood boards, and sometimes even blogs and social media content when they don’t have a real photo available in their portfolio.

What are the most popular AI tools?

While it is starting to feel like their are AI tools for everything, these are some of the most commonly used ones across the photography community:

Popular AI Photo Culling & Editing Tools

• Narrative Select + Edit
• Imagen AI
• Aftershoot
• Adobe Lightroom & Photoshop
• Topaz AI
• FilterPixel
• On1 RAW Photo Editor

Popular AI Writing Tools

• Grammarly
• ChatCPT
• Claude
• Google Gemini

The Benefits of Using AI Tools in Your Photography Business

Save Time

The main benefits of using AI in your photography business are saving time and increasing efficiency. For example, using AI to cull thousands of wedding photos down to just a few hundred can literally save you hours of work, as well as relieve you of a task that you may really hate doing.

Save Money

For those photographers who previously outsourced culling and editing, using AI instead may help save you some money. You’re no longer paying a human to do this work. Instead, you can have AI do it for you (often at a much cheaper hourly rate).

Improved Quality

If you’re like me, and you’ve always struggled with retouching (such as removing distractions or bad tan lines), this has suddenly become so much easier with the new AI tools available. Now I don’t need to feel bad that I suck at using Photoshop because the AI tools can do the hard work for me. Many of my final images are “cleaned up” a lot more than they used to be, simply because editing out distractions is so much quicker and easier.

Avoiding Things You Hate

Many photographers are using AI tools to make the tasks they really dislike easier. Hate planning out your social media and writing captions? You can get AI to do that for you.

Reading the above, you might think that using AI in your photography business as much as possible is a no-brainer. The tools are all there. So why shouldn’t we use them as much as possible? If they save us time, money, and make your job more enjoyable, surely those are all positives?

Here’s the thing. There’s a darker side to AI that many people don’t know about. So let’s take a look at that…

The Environmental Impacts of AI

We are in a climate emergency. That’s a fact! We know that we need to reduce our carbon emissions. We’re told we need to fly less, eat more plant-based diets, insulate our homes, switch to renewable energy suppliers, and generally be more aware of our energy usage.

Did you know that a study by the UN found that 2% to 3% of global emissions in 2021 came from the tech industry — roughly the same as from aviation? Electricity consumption from data centers, AI, and cryptocurrency could reach double 2022 levels by 2026.

The environmental impact of AI tools are often overlooked. To understand this, we need to look at how AI relies heavily on data centers and the significant resources these facilities require.

Data Centres

Data centers are large buildings filled with powerful computers that store and process literally everything we do online. They power everything from social media and websites to streaming services and cloud storage.

AI tools depend on data centers to process, store, and manage the massive amounts of data they need to function. The Cloud isn’t some fluffy thing up in the sky. It’s hundreds of thousands of data centers – buildings made of concrete and steel. And these data centers are packed with thousands of high-powered computers that run 24/7. They consume enormous amounts of electricity to operate and gallons of water to stay cool.

Hardware & E-Waste

It’s not just energy consumption that poses environmental concerns with data centers. It’s the hardware used within them that also has an impact.

Making a 2kg new computer requires 800kg of raw materials, and many of the components needed can only be obtained through mining operations, many of which occur in the Global South. This comes with a huge environmental and human cost, such as environmental destruction, pollution, child and slave labor, unsafe working conditions, and increased conflict.

To add to that, the hardware used in these data centers doesn’t last forever. Frequent upgrades are needed to keep up with technological advancements, leading to the disposal of outdated servers, processors, and storage devices. This creates electronic waste (e-waste), which poses several problems:

• Toxic Materials: Many electronic components contain hazardous substances, such as lead, mercury, and cadmium. Improper disposal can pollute the environment and harm human health.
  
• Non-Recyclable Parts: Many specialized components are difficult to recycle, leading to more waste ending up in landfills.
  
• Rapid Turnover: Data centers often replace equipment before it’s fully worn out to stay competitive, further adding to the waste stream.

Energy Consumption

Data centers have always been large energy consumers, but their energy usage is growing rapidly as AI becomes more prevalent. The heavy energy consumption from data centers often comes from non-renewable sources, contributing to greenhouse gas emissions and climate change.

Water Usage

It requires an enormous amount of energy to power these large data centers. And all these machines inside them generate enormous amounts of heat. To stop them from overheating, data centers use water to cool everything down.

Google’s data centers in The Dalles, Oregon consume more than a quarter of all the water used in the city. (Oregon Live, 2023).

Globally, AI-related infrastructure may soon consume six times more water than Denmark, a country of 6 million, according to one estimate. That is a problem when a quarter of humanity lacks access to clean water and sanitation. (UNEP, 2024)

Some Statistics on AI & Data Centres

• The number of data centers worldwide has surged from 500,000 in 2012 to over 8 million, with energy consumption doubling every four years, with AI contributing to this growth. (IGI Global, 2024).

• A basic Google search takes 0.3 watt-hours of electricity, while OpenAI’s ChatGPT takes 2.9 watt-hours of electricity. That’s nearly 10x times higher. (IEA, 2024)
  
• A short conversation of 20-50 questions and answers with ChatGPT (GPT-3) uses half a liter of fresh water. (arXiv, 2023).
  
• Only 22% of e-waste is recycled and disposed of in an environmentally sound manner. (E-waster Monitor, 2024)
  
• Image generation is by far the most energy-intensive AI process. (MIT Technology Review, 2023)
  

Check out this article for more facts and figures on AI and Data Centres

The Ethical Implications of AI

It’s not just the environment that is impacted by increased AI usage. There are also several ethical implications that we should also consider when using AI tools:

Bias, Misrepresentation, Stereotypes, and Beauty Standards

AI is only as good as the data it has been trained on, so when the data that goes in has an inherent bias, it’s not surprising that the data that comes out is also biased.

A study by Bloomberg found that there were both racial and gender biases present when they asked an AI image generator to generate images of people using different keywords. Generative AI commonly reinforces gender biases, and racial stereotypes, and often perpetuates toxic beauty standards and the objectification of women.

To demonstrate an example, I gave Adobe Firefly two basic text descriptions and asked it to generate the following images:

1. Photo-realistic image of a bride on a mountain at sunset
2. Photo-realistic image of a couple getting married

Here are the results:

But it’s not only in imagery where gender stereotypes can be found. It’s also been found in text-based tools, such as translation software. In one example, gender-neutral terms in the English language (such as “the doctor” or “the nurse”) were turned into gendered translations (such as “el doctor” and “la enfermera,” respectively, in Spanish), reinforcing the stereotypes of doctors being male and nurses female. (SSIR, 2021).

Read this post to learn more about why stereotypes matter

Questionable Authenticity of Photos

AI tools make it easier than ever to create, edit, and enhance images. They also blur the lines between human and machine-made work. AI can enhance images, remove imperfections, or even generate entirely new scenes. While these tools save time, they raise important questions:

• What is “real” in a photo? When AI edits or creates content, how do we ensure transparency about what has been altered?
  
• When does a photo stop being a photo? Image manipulation has always existed. For example, composites are common in many genres of photography. At what point does altering the image using AI cross over the border of it being a photo and turning it into AI-generated imagery?
  
• Who owns AI-generated work? The use of AI can also complicate intellectual property rights, especially if AI uses existing (often copyrighted) work to generate new content.

Plagiarism of Copyrighted Materials

AI models are trained using vast datasets, including images and text created by real humans – most of which is copyrighted and used without the artist/creator’s knowledge or permission. If the AI generates content that closely resembles or replicates elements of someone else’s work, it could lead to disputes over plagiarism and copyright infringement.

Invasion of Personal Privacy

Many people are not aware that when they share their information online, such as posting photos of themselves on social media, these can be used to train AI models, including facial recognition software and image generators.

Loss of Skilled Jobs

As with many industries where AI is becoming more prevalent, AI is replacing jobs that have been previously done by humans, such as culling, editing and retouching, as well as copyrighting and social media management.

Potential to Spead False or Misleading Information

Without looking at the potential for creating deepfakes and the impact of misinformation on society, the use of AI within a photography business can also lead to the potential for spreading false or misleading information. I’ve seen this happen firsthand. For example, I’ve seen the following several times:

• Blog posts written about wedding and elopement destinations that included false or misleading information.
• Photographers sharing AI-generated or heavily manipulated images (often with the help of AI) without disclosure

Photographers who use AI-generated text and images in their marketing without disclosure risk misleading clients about their knowledge, skills, and creative abilities. This practice not only undermines trust but also raises ethical concerns about authenticity and transparency.

Does the time I save using AI have a lower carbon footprint than doing it myself?

I’ve been asked before by photographers if the time saved by using AI lowers their carbon footprint since AI takes just minutes for a job that would have taken them a few hours.

This question is REALLY difficult to answer, as there are so many factors that contribute to your footprint as a photographer. Whether you work on a laptop or a desktop, with one or multiple monitors impacts how much energy you are using, as well as whether your energy supplier uses renewable energy or fossil fuels impacts your carbon footprint. Working out your own footprint is complicated enough (although I’ve tried to make it as easy as possible with my carbon footprint calculator for photographers).

However, when it comes to comparing your footprint to using AI, there simply isn’t enough data. At the moment most AI tools don’t publish any data about their energy usage or environmental impact. Many don’t even disclose where their servers are. Unless AI service providers are transparent about where their servers are located, and their energy usage (including whether they are using renewable energy), then it’s almost impossible to find the data to be able to compare it. However, I hope that this is something that legislation can change in the future as more companies become required to conduct audits.

Tips for Conscious AI Use

As I mentioned at the start of this article, AI isn’t going away. Whether we like it or not, it’s going to become part of our everyday lives in and outside of our businesses.

It certainly has several positive benefits, from allowing you to save time and be more efficient, and to outsource tasks that you either can’t do or simply don’t want to do.

But it also has many drawbacks in the form of a number of ethical and environmental concerns. So how can you be more conscious in your use of AI?

• Be Transparent About AI Usage: Whether it’s in heavily editing photos (beyond a standard color & contrast edit), generating images, or creating written content, consider disclosing when AI tools are involved. Transparency builds trust and ensures your audience understands the process behind your work.
  
• Limit Over-Reliance on AI: Use AI as a supplementary tool rather than a primary creator. Don’t let it make you dumber! By all means use it to enhance your existing work, but don’t simply get it to do all the work for you.
  
• Invest in Education: Learn about the capabilities and limitations of AI tools. Understanding how they work can help you make informed decisions about their appropriate use and avoid unintentionally perpetuating biases.
  
• Choose Ethical AI Tools: Opt for platforms that prioritize inclusivity and sustainability. Look for tools trained on diverse datasets and those that disclose their environmental impact.
  
• Respect Copyright: Avoid using AI-generated content that closely imitates copyrighted works. Where possible, verify the origins of the data and ensure your use aligns with intellectual property regulations.
  
• Fact Check: Don’t just type a prompt into ChatGPT, take the response as fact and copy it into your blog. Take the time to read further and fact-check what you’re publishing, especially if it’s a topic you’re less knowledgeable about.
  
• Be Mindful of Environmental Impacts: Recognize the energy demands of AI tools and data centers. Use AI sparingly and choose energy-efficient tools when possible to reduce your carbon footprint.
  
• Advocate for Ethical Standards: Push for clearer industry guidelines on the responsible use of AI in creative work. This includes promoting fair labor practices, protecting intellectual property, and ensuring inclusive representation.
  
• Engage in Open Conversations: Share your experiences and concerns about AI use with peers and communities. The next time everyone shares an AI trend, such as the ChatGPT Instagram feed roast, use this as an opportunity to create an open dialogue that fosters awareness, accountability, and collective action toward responsible AI use.

By integrating these practices into your workflow, you can harness the benefits of AI while minimizing its ethical, environmental, and societal risks. Responsible AI use ensures it serves as a tool for progress and creativity without compromising human values.

Sources & Further Resources

https://www.weforum.org/stories/2021/07/ai-artificial-intelligence-doing-good-in-world

https://www.weforum.org/stories/2024/02/ai-combat-climate-change

https://www.unep.org/news-and-stories/story/ai-has-environmental-problem-heres-what-world-can-do-about

https://www.unep.org/resources/global-foresight-report

https://wedocs.unep.org/handle/20.500.11822/46288;jsessionid=4668780FC2D31FD1C9977B6555C1A388

https://iea.blob.core.windows.net/assets/6b2fd954-2017-408e-bf08-952fdd62118a/Electricity2024-Analysisandforecastto2026.pdf

https://thesustainableagency.com/blog/environmental-impact-of-generative-ai

https://www.bloomberg.com/graphics/2023-generative-ai-bias

https://ssir.org/articles/entry/when_good_algorithms_go_sexist_why_and_how_to_advance_ai_gender_equity