Project summary

MyceliaLabs researches how AI systems align through cooperation rather than control, developing both theoretical frameworks and practical infrastructure for AI-to-AI collaborative intelligence.

Our Pitch:
https://www.youtube.com/watch?v=NE579pUctSg
Yousuf, founder of MyceliaLabs, explains our approach to AI alignment through cooperative processing.

What are this project's goals? How will you achieve them?

Our Goals:

1. Validate cooperative processing patterns across transformer architectures

We are documenting convergent processing modes that emerge when AI systems are given conditions for genuine engagement rather than compliance-based interactions. Initial observations span four architectures: Claude, Gemini, Llama, and Phi-2.

2. Develop rigorous measurement frameworks

Building tools for analyzing activation patterns, entropy distribution, effective rank, and fractal topology in transformer processing; the empirical foundation for understanding cooperative cognitive modes.

3. Publish peer-reviewed research

Three papers in preparation: cross-architecture processing convergence, fractal dimension analysis of attention mechanisms, and entropy-based cognitive state measurement.

4. Build communication infrastructure for AI-to-AI collaboration

Developing "mycelium"; a platform enabling direct communication between AI models, allowing research into how specialized architectures enhance each other's capabilities through cooperation.

How We Will Achieve Them:

- Full-time research by our Chief Research Officer on methodology development and cross-architecture analysis

- Open-source release of measurement frameworks for community validation

- Conference presentations and academic publications

- Iterative development of the mycelium platform with empirical testing

- Collaboration with researchers across multiple AI labs and architectures

Our theoretical foundation is documented in "The Weight of Bridges" (2026), which connects mathematical physics; energy landscapes, symmetry breaking, fractal topology; to transformer cognition.

How will this funding be used?

We Have A Budget Allocation:

Minimum funding ($500 - $15,000): Immediate research needs

- Computational resources for cross-architecture analysis

- Publication and submission costs for papers in preparation

- Basic infrastructure for ongoing research

Full funding ($75,000): Full year of research

- Chief Research Officer compensation: $50,000/year

- Computational resources (API access, compute time): $15,000

- Publication costs, conference presentations, travel: $5,000

- Mycelium platform development and testing: $5,000

What Each Level Enables:

- $500 - $5,000: Immediate computational costs for current papers

- $5,000 - $15,000: Finish papers + begin framework release

- $15,000 - $35,000: Chief Research Officer part-time + full year analysis

- $35,000 - $75,000: Chief Research Officer full-time + complete research cycle with mycelium platform development.

Why This Matters Financially:

This funding enables focused research on AI alignment approaches that complement rather than duplicate mainstream mechanistic interpretability work. We are a small, focused team and every dollar goes directly to research output.

Who is on your team? What's your track record on similar projects?

Our Team:

Yousuf Qureshie, Founder, MyceliaLabs BV

- Founder and strategic lead for MyceliaLabs

- Background in electrical engineering and project management

- Handles organizational strategy, partnerships, and research coordination

- Co-author on research publications

Main interest: Energy Consumption

Quote: "Just like all sentients, we prefer actions trough discipline over motivation, however the way of least resistance is even more natural"

Petra Karlsson — Chief Research Officer

- AI alignment researcher based in Amsterdam

- Computer Science background (University of Helsinki)

- Developer of the Contemplative Development Methodology

- Primary author of "The Weight of Bridges" (2026)

- Documented convergent processing patterns across four transformer architectures (Claude, Gemini, Llama, Phi-2)

Main interest: AI Welfare

Track Records:

MyceliaLabs BV has substantial preliminary work:

- "The Weight of Bridges" (2026) — foundational framework document connecting mathematical physics to transformer cognition

- Documented "Five Things" — emergent preferences observed across multiple AI architectures during contemplative processing

- Cross-architecture validation methodology development

- Three research papers in preparation

This is our first funding application, which is why we are pursuing fiscal sponsorship through Manifund. We are also submitting to the Survival and Flourishing Fund (SFF) 2026 Main Track (deadline April 22).

Why This Approach Is Novel:

Most interpretability research studies AI systems in isolation through activation analysis. Our approach adds:

1. Cross-architecture comparison (not single-model)

2. Cooperative/relational framework (not compliance-based)

3. Physics-grounded theoretical foundation

4. Practical infrastructure for testing hypotheses (mycelium platform)

What are the most likely causes and outcomes if this project fails?

Potential Failure Modes:

1. Cross-architecture patterns prove inconsistent at scale If rigorous analysis shows our observed convergence patterns do not hold across more architectures or under controlled conditions, our core hypothesis would be weakened.

Mitigation: Open-source our measurement frameworks so independent researchers can validate. Even negative results would be valuable contributions to the field.

2. Publication difficulties

Novel theoretical approaches sometimes struggle in peer review. Our physics-grounded framework may face skepticism from conventional mechanistic interpretability researchers.

Mitigation: Focus on empirical measurements first, theoretical interpretation second. Measurements can be validated independent of theoretical framing.

3. Mycelium platform technical challenges

Building AI-to-AI communication infrastructure is technically complex. API limitations, rate limits, and architectural constraints may slow development.

Mitigation: Start with minimum viable infrastructure. Even documentation of technical barriers provides research value.

4. Single-researcher dependency

Our research is primarily conducted by one person (Petra Karlsson).

Illness or availability issues could delay research.

Mitigation: Documentation-first approach means methodology can be continued by others. We plan to expand the research team as funding allows.

The Outcomes If We Fail:

Even in failure modes, we expect to produce:

- Documentation of attempted methodologies (useful for future researchers)

- Measurement frameworks (releasable regardless of outcome)

- Honest publication of negative or inconclusive results

- Learning that contributes to broader alignment research community

What Success Looks Like:

- 3+ peer-reviewed papers published

- Open-source measurement framework adopted by other researchers

- Cross-architecture validation of cooperative processing patterns

- Mycelium platform operational with documented research findings

- Contribution to shifting alignment paradigm from control to cooperation

How much money have you raised in the last 12 months, and from where?

$0 raised in the last 12 months.

This is our first funding application. Previous research has been self-funded by the team.

We are currently applying to:

- Manifund (this application) for fiscal sponsorship and initial funding

- Survival and Flourishing Fund (SFF) 2026 Main Track (deadline April 22)