Empirical Limitations of Generative Artificial Intelligence in Specialized Creative Domains

Abstract

This paper examines the limitations of current generative artificial intelligence systems in specialized creative applications, including fashion design, outfit recommendations, poetry generation, and short-form video. Drawing on informal observations of AI-generated outputs and interactions, it explores challenges related to interpreting visual references, understanding user intent, producing natural creative expression, and generating realistic facial appearances and synchronized speech.

The paper discusses these observations in the context of existing research on multimodal instruction following, text-to-image generation, AI-generated poetry, and facial animation. It considers the distinction between generating creative content and fulfilling the broader contextual, practical, and judgment-based requirements of human creative work.

This paper is an exploratory, practitioner-focused discussion rather than a controlled empirical study. Its observations are not intended to establish universal limitations or predict the future of artificial intelligence. Instead, it highlights why current generative AI capabilities should be evaluated carefully before claims are made about their ability to replace human creative professionals.

Integrating artificial intelligence and molecular simulation for peptide discovery: limitations, validation and decision-centered workflows
Xu Wu, Wei Fu
ChemRxiv2 Nov 2026PreprintLicence not statedGeneral chemistryChemRxiv
Bridging the Translational Gap in Artificial Intelligence for Otology and Neurotology: Clinical Readiness, Limitations, and Research Priorities
Pasquale Viola, Teodoro Aragona, Pietro De Luca +6
Medicina (Kaunas, Lithuania)30 Sep 2026Peer-reviewedCC BYFull textGeneral medicineEurope PMC
Circulating Biomarkers with Artificial Intelligence for Clinical Risk Assessment and Therapeutic Monitoring: Evidence, Limitations, and Translation
Wenyuan Wei, Rui Zhang, Zhiying Zhang
Therapeutics and Clinical Risk Management27 Oct 2026Peer-reviewedCC BY-NC 4.0Full textOncologyPMC
Artificial Intelligence in Cybersecurity: Techniques, Limitations, and Research Directions
Syed Tabish Shah
Zenodo18 Oct 2026PreprintCC BY 4.0Full textZenodo
Generative Artificial Intelligence Models in Breast Cancer Diagnosis: A Systematic Review
Diagnostics (Basel, Switzerland)1 Oct 2026Peer-reviewedCC BYFull textGeneral medicineEurope PMC

Cite

BibTeX RIS

BibTeX

@misc{dama2026empirical,
  title = {{Empirical Limitations of Generative Artificial Intelligence in Specialized Creative Domains}},
  author = {Shreya Dama},
  howpublished = {Zenodo},
  year = {2026},
  month = oct,
  doi = {10.5281/zenodo.23462051},
  url = {https://zenodo.org/records/23462051},
  note = {CC BY 4.0},
}

RIS

TY  - UNPB
TI  - Empirical Limitations of Generative Artificial Intelligence in Specialized Creative Domains
AU  - Shreya Dama
PY  - 2026
DA  - 2026/10/19
PB  - Zenodo
DO  - 10.5281/zenodo.23462051
UR  - https://zenodo.org/records/23462051
AB  - This paper examines the limitations of current generative artificial intelligence systems in specialized creative applications, including fashion design, outfit recommendations, poetry generation, and short-form video. Drawing on informal observations of AI-generated outputs and interactions, it explores challenges related to interpreting visual references, understanding user intent, producing natural creative expression, and generating realistic facial appearances and synchronized speech.  The paper discusses these observations in the context of existing research on multimodal instruction following, text-to-image generation, AI-generated poetry, and facial animation. It considers the distinction between generating creative content and fulfilling the broader contextual, practical, and judgment-based requirements of human creative work.  This paper is an exploratory, practitioner-focused discussion rather than a controlled empirical study. Its observations are not intended to establish universal limitations or predict the future of artificial intelligence. Instead, it highlights why current generative AI capabilities should be evaluated carefully before claims are made about their ability to replace human creative professionals.
ER  -

Abstract

Related

Cite