In a world where authenticity and originality are becoming increasingly scarce, understanding the hidden pitfalls of AI-generated images is crucial for anyone involved in creative work.
Whether you're a graphic designer, content creator, or simply someone interested in the evolving landscape of technology-driven art, our blog delves into the compelling reasons why you should think twice before embracing this seemingly convenient tool. Dive into our blog to uncover the truths and make sure your creativity remains genuine and untainted.
As it becomes easier to create and add AI-generated images into your documents, with tools like InDesign's built-in feature and the increased accessibility provided by Co-Pilot and Google, it's important to consider the potential drawbacks, including inauthenticity, derivative content and lack of originality. While AI-generated images may seem like an easy and convenient solution, they come with a host of problems. It's important to consider the potential drawbacks before using them in your work.
Why you shouldn't use them:
At a glance they may seem fine...
...but look carefully and they have a distinct look and often odd features like a floating chair leg, a plant that looks perfect, too many fingers, unrealistically amazing skin!
Copyright quagmire
They are derivative, a result of millions of existing images across the web being scraped. Image generators like Midjourney and Dall-E are subject to ongoing copyright lawsuits in the US and UK. You are at risk of accidentally breaking the law and infringing on someone else's copyright. There are some safer options:
Microsoft has a copyright commitment; they promise to defend customers and pay for any adverse judgments if they are sued for copyright infringement for the use of the Azure OpenAI Service outputs.
Adobe also offers customers IP indemnification for Firefly generated content.
If you are not sure, avoid using it for commercial purposes and assume copyright infringement unless you can prove otherwise.
Bias
AI-generated images can perpetuate entrenched stereotypes. The algorithms used to generate these images are trained on existing data, which can include biased and stereotypical representations of people and the world.
Environmental costs
AI-generated images also have environmental costs. The technology used to generate these images requires a lot of computing power, which can significantly impact the environment. Asking a question and getting an adequate answer from GenAI requires six to ten times the amount of power required to generate a response compared to a traditional Internet search.
The International Energy Agency estimates that by 2026, the AI industry is expected to have grown exponentially to consume at least ten times its 2023 demand.
AI-generated art can be boring
AI images are easy to spot. We have already trained ourselves to spot them. People who haven't yet will get better at it. We'll use this newfound skill to help us sift through content to help us decide if it's worth reading. They associate with inauthenticity. That's the last thing you want an evaluator to think about your proposal.
Examples...
Perhaps standards will get better over time, however, worryingly, each of the examples above is available to purchase today. AI generated images should be marked as such, so you can filter them out, or at least know to check them very, very carefully. If these wound up in your proposal, then the impression reviewers will be left with is a lack of care and attention. Not good. Be warned..
References used in this blog:
https://blogs.microsoft.com/on-the-issues/2023/09/07/copilot-copyright-commitment-ai-legal-concerns/
https://blog.adobe.com/en/publish/2024/04/16/growing-responsibly-age-of-ai-adobe-firefly-stock
https://www.forbes.com/sites/danidiplacido/2023/12/30/ai-generated-art-was-a-mistake-and-heres-why/
https://www.verdict.co.uk/the-environmental-cost-of-ai/
https://www.thomsonreuters.com/en-us/posts/esg/genai-environmental-cost/