> ## Content Index
> Fetch the complete content index at: https://kycuong.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# Joining Coefficient Giving
- URL: https://kycuong.com/joining-coefficient-giving/
- Published: 2026-11-22T02:04:58.000Z
- Updated: 2026-11-22T02:15:20.000Z
- Description: I'm a grantmaker now
- Author: Ky-Cuong Huynh

# Summary

- Last month, I joined [Coefficient Giving](https://coefficientgiving.org/?ref=kycuong.com) (CG) as a grantmaker on their [Biosecurity & Pandemic Preparedness](https://coefficientgiving.org/team/ky-cuong-huynh/?ref=kycuong.com) team.
- My portfolio focuses on AI biosecurity.
- If you're interested in working in AI safety, follow [this advice](https://80000hours.substack.com/p/how-to-get-into-ai-safety-in-3-months).
- If you're interested in learning more about biosecurity, see this [reading list](https://docs.google.com/document/d/10%5FUEhNYvGPnHMaqnVk%5Fkk4Vpugvh6d%5Fth5rZ1CMKukU/edit?pli=1&tab=t.0&ref=kycuong.com#heading=h.cir9vjmim1i9) from my colleague Chris Bakerlee.
- If you're interested in working on biosecurity, fill out my team's [expression of interest form](https://airtable.com/appk4qxUG2dzCJUT2/pagHnnCDjCWQ0tCv9/form?ref=kycuong.com).

# AI biosecurity

I'm excited to share that I've joined Coefficient Giving (CG) as a grantmaker on their Biosecurity & Pandemic Preparedness (BPP) team!

My portfolio focuses on AI biosecurity, covering questions like:

- How should we manage the dual-use capabilities of [biological AI models](https://doi.org/10.1371/journal.pcbi.1012975?ref=kycuong.com) and [LLMs](https://doi.org/10.48550/arXiv.2306.13952?ref=kycuong.com)?
- How can we best secure the dual-use [biotech ecosystem](https://doi.org/10.7249/RRA4067-1?ref=kycuong.com)?
- How should we [deal](https://www.governance.ai/analysis/coding-agents-are-changing-the-biosecurity-risk-landscape?ref=kycuong.com) with [autonomous](https://huggingface.wtf/?ref=kycuong.com) [AI](https://doi.org/10.48550/arXiv.2610.19810?ref=kycuong.com) [agents](https://www.frontiermodelforum.org/uploads/2026/07/FMF-Issue-Brief-Frontier-AI-Agents-and-Biological-Tools-Preliminary-Risks-and-Considerations.pdf?ref=kycuong.com) as a [threat actor](https://metr.org/blog/2024-11-12-rogue-replication-threat-model?ref=kycuong.com)?

These strategy documents from [BlueDot](https://blog.bluedot.org/p/biosecurity-strategy?ref=kycuong.com) and [RAND](https://doi.org/10.7249/RRA4999-1?ref=kycuong.com) give a sense of what work in this area involves.

(Strong caveats: links in this post are examples of relevant public work, not endorsements or signals of what CG funds. I do not endorse everything in these links, this post expresses only my personal views, and much of my work requires strict confidentiality.)

# What is a grantmaker?

A grantmaker is the philanthropic equivalent of a venture capitalist (VC). Our job is to make sure a problem of public interest gets solved. Usually we do so by finding and empowering people and organizations to solve that problem, because that is higher leverage than doing the work ourselves. Over the years, several grantmakers have switched to solving a key problem directly when no one else suitable was available.

Continuing the analogy: a VC is measured by the money they make for their investors. A grantmaker is measured by their [counterfactual impact](https://probablygood.org/core-concepts/counterfactual-impact/?ref=kycuong.com): the difference between what they caused to occur and what would have happened by default. (This can be negative! Obviously, my goal is for it to be highly positive.)

Usually, after I describe this analogy, people think that much of my time looks like "passive grantmaking", where you sit back and evaluate applications as they come in. But actually, much of the impact comes from "active grantmaking":

- Incubating new organizations into existence
- Bringing in people who otherwise wouldn't know about these opportunities, or who face logistical barriers
- Ideally: finding problems that no one else has identified yet ([zeroth principles thinking](https://medium.com/future-literacy/zeroth-principles-thinking-9376d0b7e7f5?ref=kycuong.com)) and then solving them

# The life of a grantmaker

In practice, my week-to-week now looks like a [manager's schedule](https://www.paulgraham.com/makersschedule.html?ref=kycuong.com), which is a big change from the maker's schedule I used to be on. Concretely, I spend my time:

- Building relationships with experts and leaders across the ecosystem to collaborate on key problems
- Deep diving into specific technical questions and running experiments as needed
- Automating my workflows (with modern models, I no longer need to write my own code)
- Reading to keep up with both AI safety and biosecurity (AI assistance is a must for prioritizing the deluge)
- Forming my own views on what is needed most
  - Perhaps no one owns a key problem today (this happens all the time)
  - Perhaps a current organization should be scaled up dramatically
  - Perhaps someone would have more impact in a different role
- Meeting with:
  - Potential grantees to discuss their plans and proposals
  - Current grantees to discuss their open questions and needs
  - Consultants to answer specialized questions
  - Contractors to coordinate specific work
- Writing grant investigation docs as part of deciding whether to fund a person or organization
- Writing strategy documents to shape action internally and externally
- Attending conferences, workshops, meetups, and other events
- Triaging inbound requests for advice and pitches for grants

This is an incomplete list, and it changes every week. I view my job as "solving catastrophic biorisk, especially from AI", and being agentic/creative/generative in how I do so is key.

It's been a wild journey to get here. I hope to write more later on "Why biosecurity?" and "Why CG?". But for now, I have a lot of work to do!