About

 

Tikanga-Ā-Kaipakihi©

Technical Support & Protection of  Hapū Data (Mātauranga Māori).

.  Hapū Data Sovereignty:  Beyond our technical skills the hapū data scientist’s success hinges on balancing innovation with ethics, ensuring fairness, and safeguarding societal trust. As the data’s influence grows across industries, the data scientist must act as steward of both truth and equity. For data science is not just solving problems—it’s about solving them responsibly.  

.  Graphical User Interface (GUI). Like your truck’s dashboard, the GUI is a visual platform (think PC, Tablet, or cell-phone USER) that allows you to interact through features like windows, icons, buttons, and menus through a keyboard to make your entries.

.  Application Programming Interface (API). This is the hinengaro (mind) where the Tikanga 2.0 elements reside to address any existential threats to mātauranga Māori.Tikanga-Ā-Kaipakihi(c) for export business focuses on equity (defined in terms of utility gains) in the Sino-Māori business where the relationship is based on Confucianism ( harmony, social order, and moral integrity),  and Taoism (focuses on balance, natural flow, and simplicity),  to complement Tikanga’s (1.0 concatenated with Tikanga-Ā-Kaipakihi(c) (that refers to the Māori way of doings things, encompassing customs, values, and social obligations to include concepts of mana, whanaungatanga, kaitiakitanga, and utu (reciprocity)).  We concede that cultural input that captures nuance will prevail over the academic theories.  We also argue that the Kalai-Smorodinsky solution has the potential to focus on proportional equity. (This is now before the Waitangi Tribunal as WAI 3248 for there is potential conflict with NZ Legislation namely the Patents Act 2013).

.  User Method (Output): (per  Williams/Tīpuna) Strategies for vetting outputs from hapū user’s PC, Tablet, and Cellular include rule-based validation. Hapū defined rules/logic checks to block output to LLMs with appropriate responses, suggestions (think spellcheck’) that violate criteria. These are point-to-point conversations hapū to the LLM. Returning responses from the LLM will be checked against external fact-checking APIs—a human-in-the-loop (HITL) (plan) review.

.  Tools & Frameworks:  A  table of approved responses will trigger corrections; fallback responses or alerts.

.  Validation libraries: Using Python’s spaCy (NLP) or great expectations (data quality) to build checks.

. Cultural Validation: with our trading partners to create allowlists/denylists for terms like whakapapa or marae (kawa, tikanga, ritenga).

. Logging/Monitoring:  Tools such as Elasticsearch© or Prometheus©  to track LLM errors, biases, or compliance issues.

. Basic LLM Response Validator. (Python code provided on request)

Conclusion: Tikanga-Ā-Kaipakihi(c) is based on clear planning, rules, and methods for the exchange of Mātauranga Māori outside of our hapū,  as a means to enforce data sovereignty, then to vet every response from the outside world from one of the many  LLMs.

 

 

About the Large Language Model (LLM). Explained in non-technical terms.

Facts: technically every LLM is trained. Think of it like training our mokopuna how to speak. She learns this way:

1. Exposure: our mōkai hears many conversations (will call that data).
2. Pattern recognition: with time she learns grammar and context (neural networks).
3. Mistakes: our kaitonotono will make errors in using the wrong word and is corrected. But remembers for the next useage.
4. Refinement: overtime our (now) rangatahi learns at Te Kohanga Reo under the expert  tutelage of the kuia or kaumātua thence to Te Kura Kaupapa Māori how to speak fluently,  but may still mirror biases heard at home (training data flows).  She goes to business, science or even law school and learns to utilise those skills by weaving tikanga expertly into Tikanga-Ā-Kaipakihi(c) into complex MLC arguments in those ‘overlapping claims and redress allocation’ applications—me ki.     

Simply put. Training an LLM mātauranga Māori means teaching it to predict, then generate text by the analysis of trillions of words from datasets, adjusting its internal mechanisms then refining its ‘understanding’ of language pattern. What about mita? How does it distinguish between the seminal works of Hirini Moko Mead or indeed PhD’s ? Maybe it can or maybe it will make a guess then call that a nuance? Anything generated by the Office of Treaty of Waitangi, or judgements/decisions from the highest courts of the land will shape  the model’s capabilities, biases, and  limitations. This makes for ethical data selection and the inclusive design critical, especially for preserving marginalised knowledge systems like mātauranga Māori.

So, any  Māori words and references present online like articles, books, fora, waiata or mōteatea or databases—would be included in the training data used to build the LLM’s data sets. So here is the challenge—the internet’s content is not exhaustive.  Does it guess? It may, it may not.  It is expected that the quality and context of data might even contain misrepresentations, or worse. This will affect the LLMs understanding and will affect how accurately the LLM interprets these terms. Also that the limitations placed on data by the various sources  will also play a large part in the interpretations.

So keep this in mind: Training data for the LLMs from  academic papers are cited in digital format like Google Books, or even in iwi owned websites/blogs that discuss culture.

The LLMs will be scanning Māori Land Court  (MOJ) findings/decisions and of course,  the TPK website.

Try not get too paranoid here. It  is expected that iwi or government and quasi-governmental Māori sources would have been vetted to ensure accuracy.  (The API should catch any misrepresentations?)

Sovereignty: Māori communities do advocate for data sovereignty—control over how their knowledge is shared and used. Keep this mind too.  LLMs will scrap culturally sensitive material without consent and that raises ethical concerns.

Homogenisation: LLMs will decontextualise knowledge leading to misunderstandings or misuse.

There are legal aspects to be considered. The LLMs does not care about UNDRIP or even Te Tiriti o Waitangi or even the other version.  

Kia Tupato (be cautious, careful, suspicious, vigilant)

Tikanga-Ā-Kaipakihi(c) cautions us all of rise of the LLM and how it presents  existential threats to mātauranga Māori rooted in culture, linguistic, and ethical challenges. These threats are around language, cultural appropriation, data sovereignty and can be exploited without recourse. TPDS teaches proactive collaboration, ethical AI practises and, the legal reforms to transform AI tools. Leadership in AI development is critical to mitigating threats and fostering respectful innovation also with our partners such as China.

 

 

Technical Specifications for the Tikanga-Ā-Kaipakihi(c) visual platform:

 

1. Frontend architecture: Vue.js for dynamic responsive UIs.

2. Backend Sevices: API gateway, authentication OAuth 2.0, JWT for user access.

3. AI/Model Interaction: JSON. Handles parameters like temperature, max_tokens for response control.

4. Security and Compliance: Data Encryption:  TLS for data in transit/encryption.

5.  Scalability & Performance: Caching: Reids/Memcached, Load balancing & CDN.

6. Monitoring & Analytics: Logging, performance metrics and User feedback.

7. Deployment: Cloud providers (AWS, GCP, Azure). CI/CD. Fallbakc Mechanisms. 8. Accessibility & Localisation: With WCAG standards & multi-language support.

 

This is an oversimplified view capturing the core elements that will be articulated in full within the Tikanga-Ā-Kaipakihi(c) documentation. 

 

Tools: 

 

Tikanga-Ā-Kaipakihi(c):  Data Abstraction is an AI tool designed for cross-cultural trade, grounded in relational, ecological, and spatial systems derived from Māori cosmology, environmental knowledge, and cultural identity. They reflect the structured, symbolic, and recursive nature of Te Ao Māori. Uncle Hek, our Pukepoto leader,  exemplified this spatial understanding. He utilised mental models of physics to calculate load distribution and structural integrity during the construction of Te Aurere. His spatial reasoning incorporated vector-like calculations, integrating direction, speed, and environmental variables such as celestial navigation, ocean currents, and manu  behavior in criss-crossing Te Moana-Nui-A-Kiwa. At TPDS, we use abstract reasoning, distinct from Western abstract computation, to develop our tools. This approach embodies a holistic form of computation, integrating the logic, geometry, and ecology inherent in Māori cultural and spiritual frameworks. This reflects a distinct epistemology where abstraction is intrinsically linked to relational and environmental contexts. Through Tikanga-Ā-Kaipakihi(c) using Kaupapa Māori we posit that Māori developed and demonstrated a form of 'living mathematics' to articulate Māori values. These values are encoded in Python using dictionaries and classes, representing each concept with its attributes. The attributes are multidimensional, encompassing relational focus, environmental connection, and spiritual aspects, each scored on a defined scale. Quantifying abstract concepts, such as the holistic framework of ecological, spiritual, and social dimensions integral to Māori ‘mathematical’ reasoning, required extensive wānanga and sensitive negotiation. A scoring system, where each attribute is rated on a scale (e.g., 1 to 5) based on its emphasis within the combined philosophy, was proposed to ensure consistency and meaningful representation. To encode Māori epistemological values alongside cross-cultural partners, we developed computational models that capture relational, contextual, and symbolic dimensions as structured vectors or semantic networks. These employ similarity metrics and allows us to computationally explore alignments between Māori relational ethics and Confucian social harmony or Taoist natural balance, while respecting each tradition's unique epistemology. Attribute and similarity thresholds are adjusted based on specific research objectives. Māori knowledge emphasises interconnectedness and sustainability, while Western sciences prioritise empirical analysis and universality. Recognising these epistemological differences fosters respectful collaboration and enriches knowledge through diverse perspectives. AI has proven effective in addressing these complexities. This kaupapa began when exporters asked how to use Tikanga Māori principles beyond Aotearoa’s shores.

 

Distinguishing Tikanga-Ā-Kaipakihi© Enterprise from others 

A CASE STUDY for Business Intelligence Researcher for Māori-China Trade

Core Concept: This initiative explores integrating Māori cultural values (Tikanga) into export strategy with China, focusing on ethical, reciprocal partnerships. A Kaupapa Māori approach investigates leveraging Tikanga (kaitiakitanga, whanaungatanga, manaakitanga) for this purpose, noting AI's potential.

Context: A hapū wānanga highlighted the need to bridge Sino-Māori beliefs. Initial analysis suggested that traditional Tikanga Māori was domestically focused and might not directly apply internationally.

Strategic Rationale: Maintaining a value-neutral global approach is unsustainable. Integrating Tikanga Māori offers a strategic advantage, similar to China's use of its cultural philosophies in global engagement.

Proposed Solution: Tikanga-Ā-Kaipakihi(c): Existing Tikanga (Mead 2003:2006) principles are being expanded into Tikanga 2.0 for cross-cultural indigenous trade. This framework aims for mutual philosophical enrichment through understanding, requiring partner acceptance based on respect and insight.

Tikanga-Ā-Kaipakihi(c) Implementation: This adaptive solution facilitates cross-cultural collaboration, emphasising convergence (not merging) despite challenges like differing worldviews and power dynamics. It requires active reciprocity and maintains distinct identities for equitable outcomes.

Key Responsibilities: This role demands diverse skills:

• Market Analysis: Researching the Chinese market (trends, regulations), analysing trade data, and evaluating potential partners.
• Cultural Intelligence: Understanding Māori and Chinese business cultures, providing briefings, and facilitating communication ("guanxi").
• Data Management: Preparing reports adhering to Māori data sovereignty principles.
• Strategic Support: Contributing to trade initiatives, collaborating with organisations, and assisting with trade missions.
• Sustainability & Ethics: Promoting sustainable practices, monitoring social/environmental issues, and researching IP protection.

Required Skills: Business degree (Master's preferred), experience in market research/international trade, strong understanding of Tikanga Māori and Chinese business culture (Mandarin ideal), analytical/communication skills, BI tool proficiency, relationship-building abilities, independence/collaboration, commitment to ethical trade, and knowledge of trade laws.

Desired Attributes: Deep understanding of Māori economic goals, existing networks, experience with indigenous communities, willingness to travel, and passion for cross-cultural relationships.

Performance Measurement: Accuracy of reports, effectiveness of briefings, successful opportunity identification, positive feedback, and tangible trade growth.

Conclusion: This Business Intelligence Researcher role is crucial for integrating cultural understanding with business intelligence to achieve successful and sustainable Māori-China trade, highlighting the specialised skills needed for nuanced cross-cultural commerce.

 

 

 

 

 

 

 

 

USER Training Syllabus:

1. Core principles of Tikanga-Ā-Kaipakihi© 

(a) Cultural Assimilation, Not imitation: Adaption & Indigenous innovation.

(b) Holistic Modernity, Integrate Tradition/Technology, Sustain Core Values.

 

2. Practical Applications

(a) Trade protocols: reciprocal trade agreements, cultural diplomacy.

(b) Digital Integration with transparency: AI and Big Data--sharing matauranga & Blockchain.

 

3. Addressing Gaps and Challenges

(a) Semantic Equivalents: expand Tikanga Maori Concepts & Cross-cultural training.

(b) Balancing Tradition and Modernity, Intergenerational collaboration & Pluralistic leaders.

 

4. Vision for the Future

(a) Global Leadership: Cultural synergy, economic growth.

(b) A Model of Holistic Modernity: create Tikanga Maori Global Contribution.

 

5. Training

Foundations of Tikanga Maori and Confucianism & Taoism. Focuses on cultural competency, business acumen, and relationship-building.

 

*  Duration: 6 months (can be adjusted based on needs).

*  Frequency: Weekly sessions (2-3 hours per session of cultural immersion experiences.)

*  Format: workshops, seminars, guest speakers, case studies, and practical exercises.

 

 

Emphasising Resilience and Adaptation

    Our Tūpuna were masters of adaptation, constantly innovating to thrive in their environment. Today, we stand at a similar crossroads with the emergence of AI. So, while political headwinds now underway seek to undermine Māori progress over the last two centuries, the global economy with our new trading partners offers new pathways. We must draw strength from our cultural foundations. Tikanga-Ā-Kaipakihi© is the embodiment of this resilience, a contemporary framework for applying timeless principles. It empowers us to navigate the complexities of AI to ensure that our cultural values guide our technological engagement. It is a declaration that Māori will not be sidelined but will actively shape our future, utilising the tools of our ancestors foretold, to forge a path of sustainable prosperity.  Te Kohanga Reo and Te Kura Kaupapa Māori graduates are our greatest source of taonga and opportunity for our future. With them, our future is bright and promising.

Embrace them!  Kia kaha!