Project summary

In January 2027, about 30 researchers, policy professionals and founders plan to spend four weeks at Hostačov, a Renaissance château in the Czech Highlands, each working on a project of their own. Most of them will pay for themselves. The money raised here would pay for those who can't, as long as their project asks how control over compute and frontier models could concentrate power, or how countries could check that rivals keep AI agreements.

František Drahota runs the house, which sleeps up to 65 people. It has six meeting spaces and a professional kitchen, with 12 hectares of parkland and forest around it. Last spring, from April to May 2026, it hosted the month-long AFFINE Superintelligence Alignment Seminar.

More about the house: https://fall.hostacov.com/

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

What I want out of this is papers, policy memos and tools that people working on AI sovereignty, power concentration and verification actually use. I'd also like some fellows to stay in these fields afterwards. They are small fields. Many researchers I mentor have promising ideas, yet never get a month of focused time, someone who reads their drafts closely or a contact in policy.

Fellows come through an open call and from the Compute Governance Career Accelerator. Each applicant submits a work sample along with a one-page plan outlining milestones. We review applications as they arrive and make offers once we secure funding. The last deadline is 13 December 2026: https://forms.gle/T4gdbK7qTQV2mBPQ9

Fellows spend most of each day on their own work. Once a week, they swap drafts in small groups and critique them. In the second week, each fellow meets with a track lead to review which milestones are on schedule; by the fourth week, everyone presents their work. An outside researcher then reviews each main output before final drafts are due on 14 February.

On 19 February 2026, the Swiss government announced that Geneva will host the next AI summit in the first half of 2027, so fellows can publish while governments prepare for it.

How will you know it worked?

Because every plan has milestones, we can see who met them. External reviews will indicate whether people in the field find the work useful. Three and six months later, I'll ask each fellow whether the project kept going, found funding or reached an organisation or policymaker who used it. I'll post what I learn here, failures included.

How will this funding be used?

Residents without funding pay the house directly: €43 per person per day when 30 residents stay, or €54 when 20 do. That includes a bed, all meals and heating. Four weeks come to about €1,200 per person.

Minimum, $17,920 (€16,000): eight fellows for four weeks

- Eight fellowship places at €1,200 each: $10,752

- Coordinator staying on site for four weeks: $3,360

- Travel bursaries of up to €150 per fellow: $1,344

- Insurance and administration, including the fiscal sponsor fee: $1,120

- Contingency: $1,344

Above the minimum, each $1,650 funds one more fellow. At $31,140, we fund 16 fellows.

At the full goal of $46,400, we fund 20 fellows for six weeks, from 4 January to 14 February 2027. With about 45 residents in the house, the price falls to about €35 per person per day.

If fewer residents come, the price per person goes up. The contingency covers that first. After that, we fund fewer fellows.

A Czech nonprofit would receive the funds as a fiscal sponsor.

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

I manage the grant and co-lead the residency with František Drahota; in November 2026, I also lead the Economy, AI & Social Change weekend at Hostačov. I have run programmes like this before. In 2025, I co-designed and led ML4Good's first AI governance bootcamp. Two years earlier, I managed research projects for the Existential Risk Alliance summer fellowship. Effective Thesis won the 2024 PACEY Award from the Basel Peace Office while I directed it. I founded 10Billion.org, whose communities in four cities, from Berlin to Jakarta, have involved more than 300 people. I mentor researchers through the Supervised Program for Alignment Research (SPAR), AI Safety Camp and Algoverse. I have held fellowships at Harvard Kennedy School and the Centre for the Governance of AI. My own research is on compute governance, geoeconomics and verification: https://doi.org/10.2139/ssrn.6361077

František Drahota runs Hostačov. His team has hosted residential programmes there through 2026, including the alignment seminar and this autumn's co-living season. He handles the venue and helps design the programme.

Jan Krasni leads the fellowship track on AI sovereignty, power concentration and verification, on site for the month. Haakon Huynh plans to join for one week and give feedback on drafts remotely. Haakon researches AI sovereignty and compute at Columbia University and writes for Tech Policy Press. With Varun Sen Bahl and me, he wrote "Sovereignty at a Price: Three AI Sovereignties and Their Tradeoffs", which he presented at the Leiden-Erasmus workshop on Digital and AI Governance. Jan leads ERIHAD.OS, an EU-funded project on federated infrastructure for sensitive research data. He works on the governance of digital infrastructures.

One paid coordinator stays on site for the four weeks. The organisers are unpaid.

https://www.linkedin.com/in/joelchristoph/

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

The biggest risk is time: we have under ten weeks to recruit. If fewer than 20 residents have committed by 18 December, we will cancel the fellowship and return the money. Fewer paying residents would also raise the daily price, so we would fund fewer fellows.

A second risk is that fellows write drafts that no organisation ever uses. That is why the application asks who would use the work. Each main output also gets an outside review.

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

For this project, $0 so far. On 26 October 2026, I applied to the GTR AI Safety Fund for €16,000 and to the Effective Altruism Funds Transformative AI Fund for €27,800. On 28 October, I applied to a foundation's capacity-building fund for €27,800. All three are pending. If any of them funds us, I will lower the goal here so the grants together never exceed the budget.

In the last 12 months, a foundation that prefers not to be named funded two of my other projects: the Compute Governance Career Accelerator and ForecastGuard, which builds AI forecasting tools for policymakers.