← Back to Blog

DAO Evolution Post-LLM: Research Findings on Tokenomics and Organizational Structure

DAO Evolution: Network nodes representing decentralized organization evolution

Executive Summary

This research document analyzes how the emergence of powerful Large Language Models (LLMs) has fundamentally altered the value proposition of Decentralized Autonomous Organizations (DAOs). Through analysis of TrueSight DAO's operational history spanning 2021-2025, we document a critical shift: cognitive work that previously required distributed human intelligence can now be performed by AI, leaving relational and physical work as the primary domain requiring human contribution.

Our findings reveal that TrueSight DAO's tokenomics system, designed with time-based and USD value-based rubrics, was already structured to handle this transition without requiring fundamental redesign. The system naturally phases out cognitive work (which AI performs instantly with zero time investment) while continuing to reward relational and physical work (which requires human time and presence).

The Paradigm Shift: Pre-LLM vs. Post-LLM DAO Operations

Before LLMs (2021-2022): The Cognitive Crowdsourcing Era

TrueSight DAO was founded in 2021, two years before LLMs became powerful enough to replace human cognitive work. During this period, the DAO structure made strategic sense because it enabled:

  • Distributed Intelligence: Multiple minds working on complex problems (legal frameworks, business analysis, research, documentation)

  • Asynchronous Collaboration: Contributors could work on cognitive tasks from anywhere, at any time

  • Specialized Knowledge Pooling: Access to diverse expertise across legal, technical, and business domains

  • Online-Only Work: All contributions could be made remotely through digital channels

During this period, tokenization focused on cognitive contributions: research (25 TDG per entry), grant applications (100 TDG per attempt), legal framework development, content writing, and business model canvas completion.

After LLMs (2023+): The Relational Work Era

By 2023, LLMs had reached sufficient capability to replace the cognitive work that DAOs were crowdsourcing. Analysis of TrueSight DAO's operational logs reveals a clear pattern:

  • AI Replaced Cognitive Work: Legal frameworks, research, documentation, and content generation are now performed by ChatGPT and similar tools in seconds

  • Human Work Shifted to Relational/Physical: Store visits, cooperative meetings, relationship building, trust establishment, and physical operations became the primary human contributions

  • Offline Work Became Critical: The work that matters now requires physical presence, face-to-face interactions, and human-to-human trust building

Evidence from operational logs shows Gary Teh and other contributors spending significant time on: visiting cacao cooperatives in Brazil, conducting facility audits (6+ hour sessions), door-to-door store visits, relationship building through dinners and cultural events, and physical logistics coordination.

Key Finding: The Tokenomics System Was Already Future-Proof

Our most significant finding is that TrueSight DAO's tokenomics system required no fundamental redesign. The rubric system is based on two metrics that are inherently resistant to AI replacement:

1. Time-Based Rewards

The rubric rewards contributions based on time spent, not activity type. This creates a natural filter:

  • AI-Performed Work: Completed in seconds → 0 hours spent → 0 tokens earned

  • Human-Relational Work: Requires hours of presence → Hours spent → Tokens earned

AI cannot spend time because it operates instantaneously. Humans must spend time to build relationships, visit locations, and establish trust. The time-based rubric automatically phases out cognitive work while continuing to reward relational work.

2. USD Value-Based Rewards

The rubric also rewards contributions based on USD value created. This further aligns incentives:

  • AI-Generated Content: May have informational value but rarely generates direct USD revenue

  • Human-Relational Work: Store visits lead to sales, cooperative relationships lead to supply contracts, trust building enables transactions → Direct USD value creation

The value-based component ensures that tokenization rewards work that creates tangible economic value, which relational and physical work inherently provides through sales, partnerships, and supply chain development.

Natural Selection Through Tokenization

One of the most elegant aspects of this system is that it requires no active management to phase out obsolete work types. The mechanism works through natural selection:

  • No New Cognitive Work Tokenized: As AI replaces cognitive tasks, no new tokenization opportunities arise for this work type

  • Members Stop Earning Tokens: Those who only contribute cognitive work stop earning new tokens (they keep existing tokens from past contributions)

  • Voting Power Stagnates: Without new token earnings, voting rights don't grow, causing natural influence shift

  • Relational Workers Gain Influence: Those doing relational/physical work continue earning tokens, gaining increasing voting power over time

  • Self-Selection: Members either adapt to relational work or naturally fade from active governance

This creates a meritocratic evolution without forced removals, drama, or hard feelings. The market (tokenization) decides what's valuable, and people self-select based on their ability to contribute what matters.

The Commitment Problem and Its Solution

Our analysis identified a fundamental structural issue with DAOs: the tension between voluntary participation (core to DAO ideals) and reliable execution (required for real-world impact).

The Core Contradiction

DAOs operate on the principle that "contribution is strictly voluntary" (as stated in TrueSight DAO's core values). However, reliable execution requires commitment, accountability, and follow-through—qualities that voluntary participation struggles to provide.

Operational logs reveal this tension repeatedly: tasks not completed on time, members getting blocked waiting on others, high churn rates, and founder doing operational work that should be delegated.

The Solution: Payment Creates Accountability

Our research suggests the solution is not to abandon DAO structure, but to separate execution from governance:

  • For Execution (Reliable Work): Find reliable people and pay them. Payment choice: Fiat (USD) for those who want money, or Tokens (TDG) for those who want voting rights

  • For Governance (Values & Vision): Maintain DAO structure with token-based voting rights

  • Hybrid Model: Execution can be paid (traditional accountability) while governance remains token-based (DAO ideals)

This approach addresses the commitment problem: payment creates accountability and reliability, while preserving the DAO's governance structure for values, vision, and long-term direction.

What Work Requires Humans Now?

Based on our analysis, the following work categories require human contribution and should be prioritized for tokenization:

Tier 1: Relationship Building (Highest Value)

  • In-person cooperative/farmer visits

  • Face-to-face strategic meetings

  • Relationship maintenance (dinners, cultural events)

  • Trust-building activities requiring physical presence

Tier 2: Physical Operations

  • Store visits with samples

  • Facility audits and inspections

  • Door-to-door sales and relationship building

  • Logistics coordination requiring physical presence

Tier 3: Network/Connector Value

  • Successful introductions that lead to deals

  • Relationship brokering

  • Trust-based facilitation

Tier 4: Reliable Execution

  • Completed physical tasks with proof

  • Follow-through on commitments

  • Accountability (showing up, delivering)

What Work Does NOT Require Humans?

The following work categories can now be performed by AI and should not be tokenized:

  • Legal Framework Research: ChatGPT can draft NDAs, licensing agreements, and bylaws

  • Grant Research: AI can compile grant opportunities and requirements

  • Documentation Writing: AI generates comprehensive documentation

  • Content Generation: AI creates marketing copy, blog posts, and social media content

  • Business Model Canvas: AI can fill out and analyze business model frameworks

  • Design Recommendations: AI suggests design approaches and tools

Exception: If someone uses AI to create something AND then executes on it physically (e.g., uses AI-generated content in a store visit), tokenize the execution part, not the AI generation part.

Implications for DAO Structure

These findings have significant implications for how DAOs should be structured:

1. DAO Structure for Governance, Not Execution

DAO structure remains valuable for:

  • Governance and voting

  • Values and vision alignment

  • Community building

  • Long-term direction setting

DAO structure is less suitable for:

  • Reliable execution (requires accountability)

  • Time-sensitive operations (requires commitment)

  • Physical operations (requires presence)

2. Hybrid Model: Best of Both Worlds

The optimal structure appears to be a hybrid model:

  • Core Operations: Traditional structure (paid employees, accountable, reliable execution)

  • Community/Governance: DAO structure (token-based voting, values preservation, ideological home)

  • Clear Separation: Execution = traditional, Philosophy = DAO

3. Payment Choice Model

For reliable execution, offer payment choice:

  • Fiat (USD): For those who want money, no voting rights

  • Tokens (TDG): For those who want voting rights, potential long-term value

  • Hybrid: Part fiat, part tokens for those who want both

This approach: attracts reliable people (payment creates commitment), separates execution from governance, solves the commitment problem, and maintains DAO structure for values.

Key Insights

  1. The tokenomics system was already future-proof: Time-based and value-based rubrics automatically phase out AI-replaceable work while continuing to reward human-required work

  2. No rubric update needed: The system doesn't care about activity type, only time spent and value created

  3. Natural selection works: Stop tokenizing cognitive work, and people doing that work naturally stop earning tokens and fade from governance

  4. Payment solves commitment: For reliable execution, find reliable people and pay them (fiat or tokens, their choice)

  5. Hybrid model is optimal: Traditional structure for execution, DAO structure for governance

Conclusion

The emergence of LLMs has fundamentally changed what work requires human contribution in DAOs. Cognitive work that previously justified DAO crowdsourcing can now be performed by AI. What remains valuable is relational and physical work: relationship building, trust establishment, physical presence, and reliable execution.

TrueSight DAO's tokenomics system, designed with time-based and value-based rubrics, was already structured to handle this transition. The system automatically phases out cognitive work (no time = no tokens) while continuing to reward relational work (time = tokens).

The key insight is that the system doesn't need to be redesigned—it needs reliable people who are paid appropriately. Payment (fiat or tokens, their choice) creates the accountability that voluntary contribution cannot provide, while preserving the DAO structure for governance, values, and long-term vision.

This research demonstrates that well-designed tokenomics systems can evolve naturally with technological change, and that the future of DAOs lies not in abandoning the structure, but in hybrid models that combine reliable execution (traditional payment) with decentralized governance (token-based voting).

Join the Discussion

We invite the community to discuss these findings in our Telegram community channel and governance platform. Share your thoughts on how LLMs have changed your contribution patterns and how we can best align tokenization with work that requires human presence and commitment.