Will the Future of Art be Artificial?
Gabriel Louis-Kayen | June 21, 2022
The use of AI-generated art has exploded in the creative industries. In the past month alone, social media has seen the wide popularization of Open AI’s Dall-E mini AI model that generates images from text— the image generator has become a popular format for creating memes. The Dall-E mini is one example of the growing field of generative AI models, using “unsupervised learning algorithms to create new digital images, video, audio, text or code” that are comprehensible and user-friendly to the often non-technical public.
*A set of images generated using the Dall-E mini model.*
While Dall-E and one of its rivals, Google’s Imagen, are text-to-image models, commonly accessible generative AI models already exist in other artistic fields. Jukebox, another Open AI project, is a neural network that generates music complete with multiple genres and coherent lines of artificially-generated singing. Within the field of literature, Sudowrite is a service that aims to curve writer’s block through generating completely cogent paragraphs.
These emergent technologies have clear potential to completely transform the creative industries as we know them. Research has already shown that individuals are unable to “distinguish and accurately identify artwork created by AI technologies” when presented with machine- and human-generated art, with up to 77% of research participants mistaking
AI-generated art for human-generated art. This leads one to wonder, will humans be replaced by machines in the future of art creation?
Current Trends
The prospect and reality of Americans losing jobs to technology are nothing new. A 2016 report by the Obama administration highlighted America’s growing dependence on AI-driven automation, and how its labor implications “might disrupt the current livelihoods of millions of Americans.” On a global scale, the McKinsey Global Institute predicts that due to automation and AI, up to “375 million workers worldwide, about 14% of the global workforce, will need to switch occupational categories by 2030 in order to avoid obsolescence.” The Brookings Institute expands on McKinsey’s data in predicting that, of these global statistics, 88 million American jobs will be affected by automation in the coming decades, with 52 million of these jobs being fairly susceptible to automation by 2030 — this equates to a third of the American labor force being impacted by the end of the decade. While these reports do not indicate specific predictions for the effects of AI on the creative industries, there is only room for an expansion in the use of generative AI in the arts. While generative AI makes up less than 1% of all current data, “by 2025, Gartner expects generative AI to account for 10% of all data produced.” Should AI-generated content begin to pervade the creative industries, what will the consequences be?
*A figure on American job susceptibility to automation created by the Brookings Institute* Implications and Consequences of AI in Art
There is no clear consensus on whether generative AI will have a positive or negative impact on the creative industries. Some argue that artificial intelligence will benefit art by reducing the barriers of entry for many artists. By making generative art easy and accessible online, the often costly labor of painting, drawing, writing, filming, etc. are removed from the creative process, allowing more individuals to participate in art creation. Additionally, many people believe that AI is yet another tool enhancing the way that an artist can express themselves by allowing them to creatively guide and constrain unsupervised learning algorithms. But, some fear that the easy development of AI-generated art will completely disrupt the creative industries and make human-based art obsolete. Other artists posit that art is defined by human activity and creativity, and that AI-generated alternatives should not even be considered art.
AI-art additionally raises a lot of concern over its ownership rights and usage. Earlier this year, former President Barack Obama delivered a speech on how AI continues to empower and worsen the effects of disinformation. The emergence of deep fakes and other synthetic media have shown the risks of AI generations being indistinguishable from authentic content. Within the creative industries, similar risks exist surrounding intellectual property (IP) rights and plagiarism. AI-art has raised a lot of unanswered questions: Will generative AI make art too easily replicable? Who falls responsible for AI-generated content that plagiarizes or steals from another individual’s work?
Analyzing trends of AI-driven art using the guidelines of the Belmont Report questions how justly generative AI will be employed within the creative industries. The Obama Administration’s 2016 report on AI and Automation indicates that AI-driven automation will disproportionately impact the jobs of lower-income and less-educated workers. The report estimates that 83% of jobs making less than $20 per hour have a high probability of automation, while only 31% of jobs making $20-40 per hour and 4% of jobs making over $40 per hour share that same high probability of automation. The report also asserts that 44% of jobs performed by those without a high school degree are highly automatable in contrast to the only 1% of jobs performed by those with bachelors degrees. While creative jobs are less automatable than most other similarly-paying jobs, lower-income artists may unjustly face the costs of automation while the financial benefits of generative art may concentrate in a small handful of individuals. Alternatively, the increasing accessibility of generative AI may make the benefits of AI-art more fairly distributed.
*AI-art generated by me using an online platform.*
Where We Are Today
It is difficult to anticipate the effects that AI will have on our society. The impacts of AI-driven automation are path dependent on how humanity’s relationship with artificial intelligence involves. Artists will need to be innovative in their applications of generative AI, but vigilant of how it detracts from the human creativity at the center of art. In a 2017 interview, Monash University Professor and artist Jon McCormack explains that AI as we know it today “is still very primitive — it doesn’t have the same capabilities as a human creative,” noting that AI models “can only draw on what they’ve been trained on.” Generative AI does not need to replace existing artistic processes, and may better serve the arts by “doing things that complement our intelligence.” AI-driven practices may open the creative industries to unforeseen innovations. Ultimately, the future of art lies in the hands of the beholder — a thoughtful partnership of artificial intelligence and human creativity will take art into uncharted yet fruitful territories.
References
- https://huggingface.co/spaces/dalle-mini/dalle-mini
- https://www.techopedia.com/definition/34633/generative-ai
- https://imagen.research.google
- https://openai.com/blog/jukebox/
- https://www.sudowrite.com
- https://www.gwern.net/docs/ai/nn/gan/2021-gangadharbatla.pdf
- https://obamawhitehouse.archives.gov/sites/whitehouse.gov/files/documents/Artificial-Intelligence-Automation-Economy.PDF
- https://www.mckinsey.com/mgi/overview/in-the-news/automation-and-the-future-of-work● https://www.brookings.edu/wp-content/uploads/2019/01/ES_2019.01_BrookingsMetro_Automation-AI_Report_Muro-Maxim-Whiton-FINAL.pdf
- https://www.gartner.com/en/newsroom/press-releases/2021-10-18-gartner-identifies-the-top strategic-technology-trends-for-2022
- https://www.cnbc.com/video/2022/04/21/former-pres-obama-takes-on-disinformation-says-it could-get-worse-with-ai.html
- https://www.hhs.gov/ohrp/regulations-and-policy/belmont-report/read-the-belmont-report/index.html
- https://app.wombo.art
● https://www.abc.net.au/news/2017-08-11/artificial-intelligence-can-ai-be-creative/8793906