Artificial intelligence + creatives: navigating with caution
In today's ever-evolving landscape, the term AI (artificial intelligence) has gained overwhelming prominence. While it sparks enthusiasm, skepticism also looms large. As a creative team, we believe exploring AI's potential with a balanced perspective is essential, acknowledging its benefits and limitations. The lines between AI, ChatGPT, and machine learning often blur, and surprisingly, some of these concepts have shaped industries longer than we might think. However, as we do our research and testing, ethical considerations arise. Within our team, opinions are divided, with some seeing AI as a valuable tool and others concerned about its implications. This post digs into the intricate and complicated relationship between AI and our creative process.
Ethical considerations
When it comes to AI’s potential, ethical questions inevitably arise. Among our team, there's a division on whether AI is a benefit or a potential threat. Three significant considerations stand out:
Consideration 1: Unintended plagiarism, lack of permission, and privacy
It's worth noting that creating these large language models or image creation involves scanning vast amounts of content, including websites, to train the AI. Our own website has likely been scanned in the process. However, no one sought our permission or asked for consent. This feels similar to all the data that Google, Meta, Apple, and others have collected from people over the past decade. We grant permission, but those vessels are no strangers to AI model training. Severe privacy concerns would arise should an AI model generate an answer that contained individually identifiable information. As creative professionals, we value proper attribution and respect for intellectual property rights. Navigating this issue requires a delicate balance between AI's capabilities and ethical considerations surrounding permission, privacy, and proper sourcing.
Consideration 2: Labor rights and effects on workers
Technological advances have always impacted labor, and often negatively. It’s the age-old people versus the machine argument. And while we tend to keep up with and embrace technology (if it aligns with our values), we are also paying attention to the negative impact AI could have on our people, specifically when considering job security and sense of belonging.
We’ve read about AI contributing to “deskilling,” inequality, and un or under-employment. And we’ve watched recent job reports communicating that AI contributed 5% of job losses in May 2023. The technology is rapidly evolving, and we think it’s crucial to keep doing our research and ensuring our people feel secure and valued in their jobs.
Consideration 3: Environmental impact and sustainability
AI also has environmental impacts, good and bad. It can help reduce the effects of the climate crisis, such as smart grid design, developing low-emission infrastructure, and modeling climate change predictions. But AI is also a major emitter of carbon. The carbon footprint of training a single big language model equals approximately 300,000 kg of carbon dioxide emissions or 125 round-trip flights between New York and Beijing. As a B Corp, we try to keep a reasonable footprint. AI is something else we will keep in mind and keep track of.
Validation through AI
Amidst these considerations, we've discovered a valuable way to incorporate AI into our creative process through validation. We do a lot of naming work — from new names for startups and products to renaming established brands. But we don't rely on AI to generate names. We have experimented with its use to validate and refine our ideas though. Here's how we approach it:
We apply AI to gauge how well our proposed names resonate. By inputting our ideas into AI models, we can measure their reception among a broader audience. This process provides insights into potential concerns, biases, cultural interpretations, and occasionally, adaptations in other languages that we might not have considered.
This application of AI complements rather than replaces our creative process, enriching it. It helps us refine our ideas and make more informed decisions. Furthermore, by harnessing AI's analytical capabilities, we ensure our creations effectively resonate in our diverse world.
Other uses
We have discovered other uses for AI in our design process. For example, we use Photoshop from time to time. Machine learning and AI in Photoshop have become an essential partner. Another example is the use of "AI frame interpolation," which has enabled us to create smoother videos. Additionally, "AI denoising" helps reduce visual noise in 3D renderings, allowing us to present sharper and more appealing designs.
Exploring the future with caution
While we're excited about the possibilities AI could offer in the future, we maintain a healthy dose of caution regarding its current outcomes. We acknowledge that we can't fully rely on AI due to its frequent errors. Nonetheless, we see its potential to improve and develop over time, perhaps eventually providing more accurate and useful solutions for our creative audience.
In the intricate intersection of AI and creativity, there's no definitive answer regarding its impact. And our team remains divided. While we currently maintain a critical view of its outcomes, we're eager to explore how it could contribute to the future of our creative audience. By striking a balance between human creativity and AI capabilities, we navigate this landscape cautiously, ensuring our creative process evolves while upholding our values and commitment to ethical practices.
