The Rise of Generative AI: Exploring its Impact on Society and Art
Artificial intelligence (AI) has become a hot topic, with some envisioning a utopian future and others fearing the downfall of humanity. However, amidst speculation about the future of AI, we must not overlook the potential impacts of the AI technologies available today.
One such technology is generative AI, which has the ability to create various forms of content such as text, images, audio, and video. Prominent generative AI systems like the chatbot ChatGPT generate conversational text by leveraging data sourced from the internet.
A commentary article published in Science by a group of 14 researchers, including individuals from MIT, sheds light on the immediate impact of generative AI on creative work and society as a whole. This article serves as a foundation for discussions surrounding this technology. Notable authors from MIT include Ziv Epstein, Matt Groh, Rob Mahari, Hope Schroeder, and Professor Alex “Sandy” Pentland.
Why was this article written?
The rapid advancements in generative AI have surpassed our previous expectations. As a result, fundamental questions about the creative process and the role of humans in creative production have arisen. Will automation replace jobs? How can we preserve the human element of creativity amidst these new technologies?
Understanding complex black-box AI systems can be challenging for researchers and the general public, hampering our ability to comprehend the societal impact of these tools. Many discussions about AI ascribe human-like intentions, agency, and self-awareness to these systems. Even the term “artificial intelligence” reinforces these beliefs. For instance, ChatGPT uses first-person pronouns and is described as “hallucinating.” By assigning agentic roles to AI, credit is undermined for the creators who contributed to the system’s outputs. Furthermore, responsibility can be evaded by developers and decision-makers when these systems cause harm.
The goal is to foster collaborations between academia and other sectors to address the immediate threats to humans posed by the deployment of generative AI tools. These threats include disinformation, job displacement, and changes to legal structures and culture.
Research Gaps in Generative AI and Art
The current discourse surrounding AI is flawed in several ways. We must expound upon how perceptions of the generative process affect attitudes towards the final outputs and the creators themselves. Additionally, it is crucial to design interfaces and systems that transparently communicate the generative process while avoiding misleading interpretations. How we talk about AI and the narratives surrounding it intersect with power dynamics. Four significant areas of concern regarding AI’s impact are aesthetics and culture, legal aspects of ownership and credit, labor economics, and the media ecosystem.
Regarding aesthetics and culture, we should take cues from past art technologies to inform our understanding of AI. When photography emerged, certain painters claimed it signified “the end of art.” However, it birthed a new medium that ultimately liberated painting from realism, leading to the rise of Impressionism and the modern art movement. Similarly, generative AI represents a distinct medium with its own unique potential. Accordingly, the nature of art will evolve alongside it. The challenge lies in how artists and creators can express their intention and style through this novel medium.
The complexities of ownership and credit present a conundrum, as current copyright laws may not adequately protect artists when AI systems are trained using their artistic styles. Determining what constitutes copying in the context of training data raises legal and technical questions. We need to ascertain whether these systems are engaging in copying and when such copying occurs.
In terms of labor economics and creative work, generative AI systems can expedite the creative process, yet they may eliminate the crucial ideation phase that originates from a blank canvas. There is value in starting from scratch, and we must investigate how AI will affect each stage of the creative process. It is paramount that we utilize these tools to complement human work rather than replace it.
Considering the impact of generative AI on the media ecosystem, we must address the risks associated with the ability to produce synthetic media at scale. There is a genuine need to protect the media ecosystem from the potential proliferation of AI-generated misinformation, striking a balance between preventing massive fraud and preserving public trust in authentic media sources.
The Reception of this Paper
This paper aims to catalyze a united and cohesive conversation about AI, which has thus far been fragmented and frustrating. Given the rapid pace of technological advancement, it has been challenging to thoroughly contemplate these ideas. To maximize the beneficial utilization of AI technologies, we must establish a shared language and determine where to focus our attention. Hopefully, this paper will serve as an initial step towards that objective. By elevating the voices of artists and other creative professionals who will be directly impacted by these tools, we can ground the discussion within the broader context of AI’s societal impact.
Artists are often at the forefront of exploring new technologies, experimenting with them long before commercial applications arise. They grapple with the ethical dimensions of these technologies and ponder their implications. The field of AI art has been active for over a decade, and artists have been grappling with the same questions that now confront society as a whole. Uplifting the perspectives of these artists and creative professionals is critical. Art serves as a fundamental expression of our humanity, occupying a central position in the broader discourse surrounding AI’s impact on society. We hope this paper can provide a foundation for that dialogue.