Architectures and Innovations in AI-Driven Humour Generation: A Comprehensive Review
DOI:
https://doi.org/10.32628/CSEIT251117237Keywords:
Artificial Intelligence (AI), Computational Humour, Large Language Models (LLMs), Multimodal Generation, Generative Adversarial Networks (GANs), Emotionally Adaptive AI, Ethical AI, Reinforcement Learning with Human Feedback (RLHF), Context-Aware Systems, Responsible Artificial Wit (RAW)Abstract
Humour, a uniquely human blend of creativity, emotion, and culture, poses a formidable challenge for artificial intelligence (AI). This review synthesizes key architectures, algorithms, and models driving AI-based humour generation—from early rule-based systems to advanced large language and multimodal frameworks. Classical humour theories of incongruity, superiority, and relief are examined in relation to contemporary techniques using Natural Language Processing (NLP), Machine Learning (ML), Generative Adversarial Networks (GANs), and Transformer architectures. Case studies of ChatGPT-4, Gemini 1.5, HumorGAN, Woebot 2.0, and Inworld AI demonstrate real-world applications across conversation, marketing, mental health, and entertainment. Ethical issues such as cultural bias, inclusivity, and creativity–coherence balance are analyzed alongside frameworks like EthicalHumorNet (2024), FunBench (2025), IEEE P7003, and OECD AI Ethics Review (2025). Looking forward, the paper envisions Responsible Artificial Wit (RAW)—AI that generates humour ethically, empathetically, and cross-culturally—marking a shift from computational imitation to socially intelligent creativity in affective computing.
Downloads
References
Anthropic. (2024). Claude 3 Model Card. Anthropic Research Publications.
Attardo, S. (2020). Linguistic Theories of Humour: Revisiting Incongruity and Script Theory. Humour – International Journal of Humour Research, 33(4), 455–472.
Binns, R. (2024). Bias, Humour and AI Ethics: Contextual Constraints in Machine-Generated Wit. AI & Society, 39(2), 541–559.
Christiano, P., et al. (2023). Fine-Tuning Language Models from Human Preferences: Updated RLHF Framework. OpenAI Research.
Devlin, J., et al. (2019). BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding. Proceedings of NAACL-HLT.
Elgammal, A., et al. (2024). HumorGAN: Adversarial Training for Creative Humour Generation. Pattern Recognition Letters, 174, 25–39.
Elgammal, A., et al. (2024). MemeGAN: Visual-Linguistic Humour Generation Using Adversarial Learning. Pattern Recognition Letters, 174, 25–39.
Fiorini, M., et al. (2025). Hybrid Neurosymbolic Models for Creative Language Understanding. Artificial Intelligence Review, 65(1), 112–134.
Gandhi, S., & Bhardwaj, S. (2024). EthicalHumorNet: Dataset and Framework for Inclusive AI Humour Generation. Journal of Artificial Intelligence in Education, 31(2), 102313.
Hernandez, F., et al. (2024). EmoLLM: Emotionally Adaptive Large Language Models for Humour and Empathy. arXiv preprint arXiv:2402.01659.
Hsu, W.-W., et al. (2024). Humour in Human–AI Interaction: Cognitive Modelling and Context Adaptation. International Journal of Human-Computer Studies, 181, 103092.
Iacobelli, A., & Melucci, M. (2024). Conversational Humour in AI: Emotional Engagement and User Trust. International Journal of Human–Computer Interaction, 40(3), 512–529.
IEEE Standards Association. (2025). P7003 Fairness of AI Humour and Satire Systems. IEEE Global Ethics Division.
Inworld AI. (2025). Character API Technical Overview. Retrieved from https://inworld.ai.
Kumar, R., et al. (2024). Affective Pedagogy and Humour Integration in AI-Based Learning Systems. Computers & Education: AI, 5, 100112.
Li, Q., et al. (2025). Creative Humour Generation for Storytelling and Dialogue Systems. Computational Linguistics, 51(2), 199–226.
Liu, T., & Lu, Y. (2024). Cultural Variability in Computational Humour and Cross-Lingual Adaptation. AI & Society, 39(3), 678–691.
Mihalcea, R., & Pulman, S. (2023). Making Machines Laugh Again: Advances in Automatic Humour Recognition and Generation. arXiv preprint arXiv:2305.07112.
OECD. (2025). AI Policy Review: Ethics and Safety in Creative Generative Models. OECD Publishing, Paris.
OpenAI. (2024). GPT-4 Technical Report. arXiv preprint arXiv:2403.XXXX.
Partnership on AI. (2024). Responsible Generative AI Framework. Retrieved from https://www.partnershiponai.org.
Rapp, A., et al. (2024). AI Humour in Marketing Communication: Cultural Context and Virality. Journal of Interactive Marketing, 51, 33–47.
Vaswani, A., et al. (2017). Attention Is All You Need. Advances in Neural Information Processing Systems (NeurIPS).
Veale, T., & Heinecke, C. G. R. (2023). Humour and Artificial Intelligence: A Critical Survey. Artificial Intelligence Review, 56(2), 457–489.
Wallace, W. (2023). Laughter as Digital Therapy: The Psychological Impact of Humour in AI Chatbots. Journal of Applied Psychology, 81(1), 91–108.
Zhou, H., et al. (2023). Multimodal Humour Generation: Combining Text, Visuals, and Videos. Neurocomputing, 557, 126612.
Zhou, H., et al. (2025). FunBench: Multilingual Humour Evaluation Benchmark for Cultural Sensitivity. Transactions on Machine Learning Research, 12(4), 1–16.
Zhou, H., et al. (2025). Visual-Linguistic Humour Fusion in Gemini 1.5. Neurocomputing, 591, 126943.
Downloads
Published
Issue
Section
License
Copyright (c) 2025 International Journal of Scientific Research in Computer Science, Engineering and Information Technology
This work is licensed under a Creative Commons Attribution 4.0 International License.
