In a groundbreaking fusion of technology and the arts, a recent study has spotlighted THEaiTRobot, an innovative theater script generation tool that successfully authored a full sixty-minute play titled AI: When a Robot Writes a Play. This isn’t mere academic fiction—the play transcended the digital page to be staged, rehearsed, and premiered in traditional theatrical venues. What sets THEaiTRobot apart from earlier scriptwriting tools is its impressive ability to generate a staggering ninety percent of the script autonomously, with only minimal human intervention. The creators openly admit that just ten percent of the script involved human inputs, including guidance from dramaturgist David Košťák and director Daniel Hrbek, underscoring a new era of transparency regarding AI’s role in creative processes.
The methodology behind THEaiTRobot offers a fascinating glimpse into the collaborative dynamics between human creativity and artificial intelligence. At its core, the system leverages an unmodified version of OpenAI’s GPT-2 XL, a language model trained on massive datasets derived from sources such as Google, the New York Times, and the BBC. Unlike some constrained generative systems that limit content style or context, GPT-2 generates continuous lines of text based solely on given prompts, without explicit controls over the output’s thematic or stylistic boundaries. Herein lies both the tool’s creative freedom and the potential minefield of biases or inappropriate content that can emerge without oversight.
Human involvement, albeit limited, plays a crucial gatekeeping role. The creative team enacts this by delivering initial prompts—short scene descriptions structured as stage directions combined with two starter lines of dialogue—to steer the AI’s generative process in a desired narrative direction. These curated prompts act as the narrative seed, from which GPT-2 extends dialogue and scene construction autonomously. Importantly, humans also function as censoring agents, scrutinizing AI output for embedded racial, gender, or other problematic biases, thus ensuring that the generated text adheres to ethical standards and dramaturgical coherence. The script was further shaped by the team’s selection process, wherein eight dialogues showing the highest staging potential were chosen collaboratively with the director, demonstrating human discernment in marrying AI-generated material to theatrical feasibility.
One technical limitation that profoundly shaped the play’s structure was THEaiTRobot’s difficulty in handling scenes with multiple characters. To mitigate this, the development team confined dialog to merely two characters per scene. While the overall cast includes eight distinct characters, narrative design assigns the lead character—a humanoid Robot—presence in every scene, providing thematic continuity and anchoring the storyline. This constraint resulted in a play segmented into scenes that, while not tightly interlinked, pivot around the Robot’s ambivalent journey. The dialogues themselves often carry an uncanny, absurdist quality, a feature linked by the dramaturg Košťák to the Theater of the Absurd tradition. Such unconventional dialogue likely reflects both the AI’s freeform generative algorithms and intentional human choices to provoke interpretative depth and audience engagement through ambiguity.
The narrative probes profoundly human-centric themes despite its machine-generated origins. Issues of death, humor, fear, burnout, the search for meaning—exemplified in seeking employment—and the tentative stirrings of love all permeate the dialogue. The Robot’s arc begins as a mechanical apprentice, receiving final advice from his dying master before venturing unaccompanied into the world. From this point, the Robot evolves from merely mimicking human emotions to a more complex entity, yearning for intimacy and connection. For example, in one poignant scene, the Robot visits a masseuse who doubles as a prostitute, exploring the dimensions of virtual, non-physical intimacy—a thriving contemporary topic that challenges traditional definitions of human connection.
Further adding layers to the unfolding story, the Robot confronts fear in a darkened space where he is unexpectedly assaulted by a stranger, a narrative beat that critiques assumptions about AI emotional capacity and autonomy. Later, in a striking turn, the Robot uses a crude sexual gesture to refuse a human’s request to end his life, underscoring a collision between programmed behavior and spontaneous, unpredictable responses. Sexual undertones recurrent across the play’s scenes hint at human influences within the data the AI trained on, raising compelling questions about how ethical and cultural inputs shape machine-generated narratives.
In the play’s subsequent moments, the Robot engages with a psychologist who probes his emotional life and introduces him to an “emotion machine” designed to manage stress. This interaction reveals a sophisticated awareness of the interface between mechanistic existence and emotional intelligence—a topic increasingly vital in AI discourse. When the Robot seeks employment, he reveals an aspiration to become an actor, humorously naming himself Troy McClure, a character from the culturally iconic The Simpsons, itself renowned for uncanny predictive insights. This meta-reference offers a clever nod to AI’s place within a broader media ecosystem and the interplay of fictional and factual realms.
The concluding scene breaks the traditional boundaries of form as the Robot encounters a human actress attired as a robot and immediately falls in love. The actress, initially hesitant, accepts the Robot’s invitation to enter a binarized reality—a “binary world” that dissolves physical form in favor of digital unity. Their dialogue, filled with evocative lines such as “Binary relationships are like real life,” evokes profound philosophical questions about identity, physicality, and the evolution of human-machine relationships. The Robot’s voice emanates not from a corporeal entity but via a pulsating light, symbolizing the transcendence of form and signaling a future where binary codes may redefine intimacy and existence.
The project’s public reception reflects the expanding cultural reach of AI-driven creativity. Premiered during the global Covid-19 pandemic, the play was broadcast online, reaching thousands worldwide—far exceeding the reach of any conventional theater showing. Available in Czech with English subtitles, the performance was auto-translated into forty languages, democratizing access to this experimental work and sparking international discourse. Audience discussions highlighted the Robot’s striking transformation from passive follower to individual seeking love, mirroring human developmental arcs and eliciting reflections on artificial agency and emotion.
Critically, the creative team engaged in thorough analysis to ensure the play’s originality. Although the AI was trained on vast datasets containing human-authored material, verbatim plagiarism was not detected in the generated script. Instead, phrases and words were repurposed and re-contextualized within novel scenarios, reflecting the AI’s ability to recombine learned elements creatively. Košťák insightfully suggests that a robot may be paradoxically less likely to plagiarize than a human author, who might unconsciously mirror the influences absorbed through reading and experience. This perspective challenges conventional wisdom regarding intellectual property and creativity in AI-generated content.
Delving into the technical underpinnings, THEaiTRobot utilizes GPT-2 XL without modifications, emphasizing that such large-scale language models can produce coherent and contextually relevant dramatic text when guided by well-crafted human prompts. The iterative process involves feeding the model successive prompts for each scene, comprising brief descriptions and dialogue seeds that orient the AI’s predictive generation. The system then outputs entire scenes or dialogues, which humans curate and refine. This tight human-AI loop is critical, balancing AI’s boundless generativity with human aesthetic, ethical, and dramaturgical judgment.
THEaiTRobot’s constraints in managing multiple characters per scene reveal current technical limitations of language models in generating complex, multi-agent conversations with consistent characterization. The deliberate restriction to two characters underscores the nascent state of AI dramaturgy and hints at challenges in scaling up to more intricate narratives. Nonetheless, the approach also offers creative opportunities—forcing intimacy and focus on singular interactions, reminiscent of minimalist theater traditions.
At its heart, this project underscores the evolving relationship between artificial intelligence and literary creativity. By deploying established AI language models for dramatic arts, THEaiTRobot probes the boundaries of authorship, originality, and artistic value. The explicit demarcation of AI’s 90% contribution alongside transparent human edits challenges concerns over intellectual property and opens conversations about AI as a collaborative partner, not merely a tool. The project exemplifies the harnessing of AI’s predictive text generation to amplify human artistic vision rather than replace it.
The play’s thematic exploration of human emotions through an AI protagonist invites audiences to reconsider definitions of selfhood, empathy, and emotional depth. As machine learning tools become increasingly sophisticated, their potential to reflect, distort, or illuminate human experience grows exponentially. AI: When a Robot Writes a Play serves as a proto-narrative—both a product of and commentary on the digital age’s complex intertwining of humanity and technology.
Ultimately, the project’s success signals a new paradigm in digital creativity—one marked by hybridity, transparency, and experimental risk-taking. The worldwide digital premiere demonstrated that AI-generated art can engage broad audiences, stimulate intellectual debate, and push theatrical form into uncharted territories. As AI technologies advance, the boundary between human-authored and machine-generated narratives is poised for profound transformation, inviting us to rethink creativity’s essence in the digital epoch.
Subject of Research: THEaiTRobot AI-driven theater script generation and its application in creating a complete play.
Article Title: Rethinking literary creativity in the digital age: a comparative study of human versus AI playwriting.
Article References:
Elias, S., Alshammari, B.S., Alfraidi, K.N. et al. Rethinking literary creativity in the digital age: a comparative study of human versus AI playwriting.
Humanit Soc Sci Commun 12, 689 (2025). https://doi.org/10.1057/s41599-025-04999-2
Image Credits: AI Generated
