Page 55 - DP-2-2

Design+ Da Vinci AI Tutor in art history learning

et al., which allow natural language interactions to a sense of presence and engagement that is often missing in
37
enhance user engagement with cultural exhibits. traditional instruction.
These initial efforts underscore the potential of In recent years, advancements in AI have substantially
generative tools in making history accessible and transformed educational environments, particularly
participatory. Recent technological advancements have through generative platforms capable of creating
also taken this concept further by blending smart systems personalized and interactive learning experiences. While
with immersive environments. For example, AI Time considerable research has demonstrated the potential
Machine (https://www.myheritage.com/ai-time-machine) of the tutors within STEM education, such innovations
by MyHeritage uses Stable Diffusion to create personalized, have only recently begun to penetrate the humanities
historically themed photo transformations, linking disciplines, resulting in significant knowledge gaps
the modern identities of users with historical periods. regarding their effectiveness in nuanced and context-rich
Similarly, immersive VR experiences, such as Virtual environments. Emerging research has defined desirable
Angkor (https://www.virtualangkor.com/) by Monash characteristics and necessary conditions for AI teaching
University, enable dynamic interactions with reconstructed assistants, underscoring the importance of contextual
historical settings, providing a deeper cultural and social adaptability, personalized feedback, conversational
understanding of past civilizations. In fact, AI-driven coherence, and empathetic communication to enhance
38
tools are increasingly blurring the lines between passive learner engagement. In particular, Denny et al. argue
42
observation and active engagement in historical learning. that effective intelligent tutors must provide adaptive
Applications such as the Hello History AI (https://www. responses, foster dynamic interactions, and maintain
hellohistory.ai/) simulate meaningful conversations a conversational tone that is conducive to sustained
with over 20,000 historical personas, demonstrating the learner engagement – qualities that are vital in humanities
feasibility of neural networks in capturing both knowledge education.
and personality. Other projects, such as the Chinatown
39
AR Project (https://polymathic.usc.edu/ahmanson-lab/ In addition, recent comparative studies by Hajeer
43
chinatown-history-project) and the VR reconstruction of et al. exploring perceptions of these tutors versus human
the Vatican Stanza della Segnatura (https://polymathic.usc. instructors among higher education students indicate both
edu/ahmanson-lab/building-virtual-renaissance-library- enthusiasm and caution in student attitudes. Although
julius-iis-stanza-della-segnatura) by the University of students recognize the strengths of such automated
Southern California Dornsife, demonstrate how augmented tutors in promoting autonomy, facilitating self-paced
and VR tools can democratize access to historical contexts. learning, and ensuring immediate and adaptive feedback,
These innovations illustrate how generative AI and XR significant limitations persist in the realms of emotional
technologies are creating unprecedented opportunities intelligence, subjective interpretation, and interpersonal
for personalized, interactive, and engaging historical communication nuances. These limitations are particularly
experiences. acute in disciplines such as art history, where nuanced
While projects using generative interfaces have made interpretation and emotional connections with subject
significant strides in historical engagement, the integration matter play crucial pedagogical roles. Consequently, their
of multiple technologies – such as LLMs trained on findings suggest a complementary relationship between
discipline-specific datasets, XR avatars, and historical digital tutors and human educators rather than one of
figures – presents a new frontier in education. LLMs, direct substitution. Furthermore, immersive technologies,
including those tailored for educational applications, when integrated with intelligent tutoring systems, enhance
have demonstrated remarkable potential in delivering the depth and authenticity of learning experiences within
44
personalized learning experiences, adapting dynamically to humanities education. Gao et al., in their examination
individual student needs. These systems can contextualize of digital avatars within 3D virtual reconstructions,
40
historical knowledge in a manner that engages students demonstrate how such blended platforms significantly
while addressing specific gaps in their understanding. For boost historical empathy, learner engagement, and
instance, recent work highlights how retrieval-augmented contextual comprehension through realistic simulations
generation techniques can tailor LLM responses to align of historical environments, as exemplified in their case
closely with structured course materials, providing detailed study of the Temple of Demeter in Greece. Integrating
and accurate feedback to student queries. Extending this historically accurate avatars within rich spatial contexts
41
model, the integration of XR avatars allows for the visual allows learners to interact directly with historical figures
and behavioral embodiment of historical figures, enabling and locations, fostering not only deeper understanding but
not only deeper interaction with content but also fostering also stronger emotional connections to the subject matter.

Volume 2 Issue 2 (2025) 5 doi: 10.36922/dp.8365