Proposed by 3Dkosmos
HPCNepal (hpcnepal.org) is a Kathmandu-based not-for-profit organization founded in 2019 with a deceptively simple but ambitious question at its core: How can we create the High Performance Computing resources that a country needs?The organization is visioned to create a scientific research culture using HPC — to foster scientific innovations and to help universities and research institutions acquire world-class capacity in high-performance computing, ultimately building next-generation expertise and workforce for a knowledge-based, smart society and economy. The inclusiveness caters to hospitality and tourism sector .
Description
Generative AI threatens to homogenize cultural expression. Most AI models are trained on Western-centric data, producing outputs that flatten or misrepresent non-Western cultural norms, etiquette, and heritage practices. For the hospitality industry — a primary gateway for cross-cultural encounter — this creates reputational and financial risk. Hotels serving international guests increasingly rely on AI guidance, but existing solutions lack transparency and cultural accuracy. GuestSense AI is a transparent, auditable generative AI platform that preserves and accurately applies cultural heritage knowledge in real-time hospitality settings. Unlike black-box models, GuestSense AI sources its cultural intelligence from curated, verified heritage datasets — including academic ethnographies, UNESCO cultural practice descriptions, and community-reviewed etiquette guides from 50+ nationalities. Key innovations aligned with the call: Provenance transparency – Every cultural recommendation (e.g., "bow, don't shake hands for Japanese guests") includes a traceable source citation and confidence score. Users can audit why the AI made a specific suggestion. Fair compensation framework – GuestSense AI proposes a micro-licensing model where cultural knowledge contributors (anthropologists, local heritage organizations, cultural attachés) receive compensation each time their data informs an AI-generated recommendation. This creates a fair market for cultural heritage content in the AI era. Anti-homogenization architecture – The system deliberately preserves cultural nuance and diversity. It flags when cultural practices conflict (e.g., Chinese vs. Indian breakfast expectations) rather than forcing a "neutral" Western solution. Community validation loops – Cultural heritage organizations review and correct AI outputs, creating a continuous feedback mechanism that prevents model drift toward dominant cultural norms.