AI Diplomacy: Singapore Foreign Minister Uses Agentic AI As Second Brain
Singapore’s Foreign Minister Dr. Vivian Balakrishnan recently made waves not with a new treaty or a diplomatic speech, but with something far more personal: his own AI “second brain.” Built on a Raspberry Pi using open‑source tools like NanoClaw and the LLM Wiki pattern, this system is reshaping how high‑stakes professionals capture, connect, and reuse knowledge. For anyone serious about their career—especially in fast‑moving domains like tech, policy, or international business—this is not just a curiosity. It’s a blueprint for a new kind of professional edge.
What is a “second brain”?
At its core, a second brain is an external system that stores your ideas, notes, decisions, communications, and context so you can retrieve them exactly when you need them. Instead of relying only on memory and scattered folders, you create a structured, searchable, and evolving knowledge base. Balakrishnan’s version goes further: it’s an intelligent second brain that reads your messages, extracts facts, and connects them into a living “wiki” that can answer complex questions about your own work and history.
In practice, this means:
- Every email, briefing, or voice note can be turned into structured knowledge.
- Past decisions, talking points, and stakeholder positions are instantly retrievable.
- The system gets “smarter” the longer you use it, because it learns which facts and connections matter to you.
How Balakrishnan’s setup works
Balakrishnan’s assistant, NanoClaw, runs on a Raspberry Pi 5 hosted on his own network. This local‑first architecture means sensitive diplomatic and policy content never leaves his control. NanoClaw connects to his messaging channels (WhatsApp, Gmail, etc.) and feeds everything into an internal knowledge graph instead of treating each message as a one‑off event.
The key breakthrough is the LLM Wiki pattern. Rather than simply doing keyword search over a pile of documents, the system:
- Extracts facts from raw text (e.g., “Country X opposes measure Y due to concern Z”).
- Stores them as structured entries in a SQLite‑backed knowledge store using tools like mnemon.
- Connects related entries (entities, dates, issues, positions) so questions can be answered by traversing relationships, not just text chunks.
This graph then sits under a front‑end powered by Obsidian or similar tools, giving him a visual “wiki” of his own work. When he asks the system, “What was our position on topic T in 2023?” it doesn’t just find a single document—it reconstructs the relevant context from multiple sources, linked together.
What this can do for you
Even if you aren’t a foreign minister, a second‑brain system like this can transform how you work:
1. Defense of your time and reputation
A second brain preserves your track record. Every project decision, negotiation, or stakeholder position is logged in a way that can be recalled months or years later. If you’re asked to justify a past choice or orientation, the system surfaces the evidence, not just your memory. This is especially powerful in complex, cross‑border roles where subtle shifts in position or context matter.
2. Faster, more confident expertise
Instead of Googling everything or re‑reading old emails, you ask your brain: “What are our past assumptions on X?” or “What risks did we flag in the last feasibility review?” The system returns a compact, curated answer with links to raw material. Over time, you develop a personal ontology—your own taxonomy of issues, players, and patterns—that makes you faster and more consistent.
3. Serendipity and pattern‑finding
Raw documents are hard to mine for patterns on demand. A knowledge‑graph‑based second brain, however, can show you clusters of topics, recurring stakeholders, or repeated risks. You might discover, for example, that a particular vendor class always raises compliance issues, or that a certain region surfaces the same regulatory bottleneck. This turns reactive note‑taking into active pattern‑finding.
4. Safer, more private workflows
Because Balakrishnan runs his assistant on a Raspberry Pi on his own network, he avoids the security and privacy risks of sending sensitive content into a third‑party cloud LLM. For professionals handling confidential business, policy, or technical data, this privacy‑first model is the gold standard. You still get the power of large language models; you just keep the data where you control it.
5. Automation without “magic”
The system doesn’t just answer questions. It can automate reminders, draft follow‑ups, or surface relevant precedents when you start a new email or brief. But because it’s built on open‑source tools and clear architecture, you remain in control. You know exactly how it works, where the data lives, and what it can and cannot do.
Why this matters for your career
For someone mapping a path into New Zealand’s tech, policy, or international sectors, a second brain is a powerful way to:
- Stand out in interviews by showing a structured, documented track record of projects and decisions.
- Accelerate on‑boarding by quickly accessing your own previous learnings about vendors, regulations, and stakeholder maps.
- Manage complexity across time zones, partners, and jurisdictions, without losing nuance.
Balakrishnan’s implementation is a reminder that elite performance isn’t just about smarter tools; it’s about self‑hosted intelligence that reflects your own thinking over time. The more you feed it, the more it becomes an extension of your professional identity.
Ready to build your own second brain?
If you’re serious about building a system that preserves your knowledge, accelerates your decisions, and keeps your data private, you don’t need to start from scratch. You can adopt patterns like NanoClaw, LLM Wiki, and mnemon, then tailor them to your workflows in tech, procurement, or international business.
If you’d like help designing a second‑brain architecture that fits your goals—whether you’re a high executive, farmer, global policy roles, or cross‑border digital‑transformation projects—reach out to us.
Contact us at [email protected] to start the conversation.