
OUR HUMAN-AI INTERACTION
(HAI) RESEARCH
some of
OUR PAST WORK
-
Comparative Analysis of GPT-40 and Gemini 1.5 Pro in Thai Exam Settings
This comprehensive work offers valuable insights into the performance of artificial intelligence in standardized Thai examinations, including the prestigious POSN Biology, Mathematics, as well as A-Level Thai Language and Social Studies. The research meticulously highlights the comparative strengths and weaknesses of GPT-40 and Gemini 1.5 Pro, particularly in terms of accuracy and efficiency. Notably, it reveals that GPT-40 excels significantly in analytical subjects, showcasing superior capabilities, while Gemini 1.5 Pro demonstrates a remarkable strength in language comprehension tasks, effectively illustrating the differing areas of expertise between the two AI models.
-
SIR Model Analysis of Future Epidemics Diseases "PathogenX" Using AI
Our project utilizes the SIR model to comprehensively study potential future diseases, such as the hypothetical "PathogenX." By integrating advanced AI technologies, we aim to significantly improve the accuracy and reliability of epidemic predictions. The primary objective of our initiative is to enhance the traditional SIR model with AI capabilities, enabling us to better forecast the spread and impact of emerging infectious diseases. We will delve into the behaviors and characteristics of potential future epidemics, thoroughly explore various transmission dynamics and recovery rates, and ultimately provide valuable insights that can enhance public health readiness and effective response strategies.
-
AI Harmony: Integrating Hormonal Mechanisms into LLMs for Improved Decision-Making
The study has shown exceptionally promising results, with the innovative Hormone-inspired GPT model, aptly named Monkan Kankoon, significantly outperforming the standard GPT 4 in a variety of cognitive and emotional tasks. Notable and significant improvements were observed in exam performances, decision-making accuracy, and emotional regulation capabilities. This advancement could potentially pave the way for the development of more human-like AI systems that can better understand, interpret, and respond to the intricate complexities of human emotions and diverse needs.
-
The Analysis of the SIR Model to Study Different Patterns of Future Outbreaks of a Dangerous Disease (Pathogen X) Using AI
This study leverages the SIR model enhanced by AI, particularly GPT-4, to unravel how variations in transmission rate (β) and recovery rate (γ) can shape the trajectory of deadly outbreaks. The research delivers critical insights through four distinct scenarios, showcasing actionable strategies to mitigate peak infections, reduce healthcare burdens, and shorten outbreak durations.
-
Sensory Construction of Dreams in Blind Individuals: A Comparative Analysis by Vision History, Age, and Sex
Blind individuals construct dream experiences using the senses available to them—we reviewed and synthesized data from peer-reviewed studies comparing congenitally blind, early-blind, and late-blind groups. Aggregating weighted means for visual, auditory, tactile, olfactory, gustatory imagery, and nightmare frequency (ANOVA/Tukey HSD; p<0.05), we found that: Congenitally blind dream without visual imagery but report very high auditory/tactile sensations and a doubled nightmare rate (~25%). Early-blind show limited, fragmentary visuals (~20%) alongside non-visual senses. Late-blind retain rich visual dreams (~80%) that gradually diminish over time. For age and sex had negligible effects on these sensory patterns, underscoring the brain’s adaptability in dream construction.
-
Sensing the Soul: A Biometric-Exoskeleton Model for Visualizing Human Chakra Dynamics
This proof-of-concept project bridges ancient chakra systems with contemporary physiological science and wearable technology. By mapping each of the seven chakras to specific biometric signals—such as EEG, EMG, HRV, respiratory and postural data—the study demonstrates how spiritual traditions can be quantified and visualized through modern digital sensors and exoskeletal interfaces. The research features comprehensive tables and visualizations that highlight physiological markers for chakra balance and imbalance, laying groundwork for future applications in holistic well-being, biofeedback, and integrative mind-body technology.