Aletheia Project Proposal – Foresight Institute Submission
Kayode Adekoya - Economist | AI Safety Advocate
June 2025.
1. Executive Summary
Aletheia is an open-source AI assistant designed to automate and enhance literature reviews in safety-critical domains like AI safety, biosecurity, and longtermism. It integrates semantic search, summarization, and citation mapping into one platform, helping researchers synthesize knowledge more quickly and accurately. Aletheia emphasizes transparency, human oversight, and responsible AI deployment.
2. Project Overview
This project will develop Aletheia, a modular and user-friendly AI tool aimed at academic researchers and forecasters. Its features include literature synthesis, concept mapping, and summarization using natural language processing. Tailored to fast-evolving fields like AI safety, it prioritizes ethical usage and scientific rigor.
3. Problem Statement
Researchers in fields like AI alignment face a massive and growing volume of literature. Traditional literature reviews are labor-intensive and error-prone, causing delays in research and policy development. This slows progress on critical long-term challenges.
4. Current Limitations in the Field
Existing tools such as Google Scholar, Elicit, and Semantic Scholar are helpful but fragmented. They often lack full summarization capabilities, semantic context, transparent sourcing, or features tailored to safety-critical domains. Furthermore, most are proprietary and not open to community contributions or scrutiny.
5. Proposed Solution: Aletheia
Aletheia will combine the strengths of multiple tools in a single open-source platform that:
- Uses AI to extract and summarize key insights from academic texts
- Maps citations and intellectual lineages visually
- Supports semantic search with human-in-the-loop design
- Promotes transparent, ethical use aligned with AI safety principles
6. What’s New & Why It Will Succeed
Aletheia’s novelty lies in its integration of summarization, search, and citation mapping for safety-critical research. It uniquely combines transparency, user governance, and ethical guardrails. These features directly address researcher pain points, and the tool’s modularity supports community-driven improvement.
7. Success Criteria
Success means:
- Literature reviews that once took weeks now take hours
- Widespread adoption by AI safety researchers
- Citations in published papers and policy documents
- Verified improvements in researcher productivity and insight quality
- Community engagement through forks, contributions, and feedback
8. Risk Assessment & Mitigation
Risks:
- Misinterpretation of AI summaries
- Spread of low-quality or biased data
- Over-reliance on automation
Mitigations:
- Human-in-the-loop review before conclusions are used
- Transparent metadata and audit trails
- Academic peer-review features for flagged content
9. Timeline & Milestones
Start Date: August 1, 2025
End Date: February 28, 2026
Milestone 1: Research Audit (Aug 2025)
Milestone 2: MVP Prototype – Search + Summarization (Oct 2025)
Milestone 3: Citation Mapping Module (Nov 2025)
Milestone 4: Beta Testing & Community Feedback (Jan 2026)
Milestone 5: Full Release & Documentation (Feb 2026)
10. Budget & Cost Breakdown
Total Budget: $62,500
- AI Developer (6 mo): $30,000
- Research Assistant: $10,000
- UI/UX Design: $5,000
- Hosting & Infrastructure: $3,000
- APIs & Tools: $2,500
- Outreach & Testing: $5,000
- Licensing & Legal: $1,000
- Miscellaneous: $6,000
11. Scaling Plan (3–5 Years)
Post-launch Aletheia will:
- Add multilingual and domain-specific models
- Integrate with forecasting tools and policy dashboards
- Collaborate with Effective Altruism and AI safety communities
- Offer plug-ins for specialized use cases (e.g., biosecurity, ethics)
- Establish sustainable open-source governance
12. Track Record / Founder Background
Kayode Adekoya is an economist, AI safety advocate, and founder of The AI Guardian. He has led research, advocacy, and public engagement initiatives focused on AI governance and responsible development. His experience combines public sector analysis and community-based awareness, positioning him to lead Aletheia effectively.
13. References & Academic Basis
Selected References:
- Fasanya & Fasanya (2023), 'Harnessing AI-Powered Search Engines for Literature Review' [Link](https://digitalcommons.unl.edu/libphilprac/8043/)
- Khalifa & Albadawy (2024), 'Using AI in Academic Writing and Research' [Link](https://www.sciencedirect.com/science/article/pii/S2666990024000120)
- PhilPapers, Semantic Scholar, and Elicit.ai tools
- Alignment with Foresight Institute’s RFP on Automating Research & Forecasting
