The latest comments and grant applications on grantmaking.ai.
The latest comments and grant applications on grantmaking.ai.
Applied to grantmaking.ai Launch Round · $8,000 requested
Lorem ipsum dolor sit amet, consectetur adipiscing elit. Praesent iaculis mollis mi, non vestibulum quam sollicitudin a. Curabitur quis maximus eros. Ut vitae posuere nulla, vel congue ligula. Sed quis turpis orci. Sed nec gravida urna, sit amet semper sem. Aenean vehicula accumsan est, vel posuere felis sagittis ac. Nulla aliquet aliquet felis ac commodo. In posuere vel velit at eleifend.…
Hey, do you know if Leo closed his microgrant program? I got removed from the category, and when I added it back, it got removed again. I tried asking Austin last night and have yet to receive a response.
@nicholas Nevermind. The logs have just been updated; he deposited $50,000 USD more into his account.
@NeelNanda — You funded the Asterisk AI Blogging Fellowship because good writing about AI reaches people that research papers don't. This project applies the same logic to Traditional Chinese.
Taiwan AI Governance Forum is the only high-quality AI governance publication in Traditional Chinese serving Taiwan's manufacturing sector — the workforce that builds the hardware on which frontier AI…
@lisathiergart -Your focus on hardware security and technical governance maps directly onto this project.
The EU Cyber Resilience Act requires cybersecurity conformity assessment for all connected devices sold into Europe, with high-risk obligations applying August 2026. Taiwan manufactures the majority of these devices. The engineers and compliance teams responsible for CRA conformity read…
@MarcusAbramovitch — You just funded Tarbell for AI journalism. This project is the same category for a geography nobody else is covering.
Taiwan produces the majority of the world's AI chips. Its factory owners and senior managers are making real-time decisions about AI deployment with zero access to governance information in their language. This is not a meta-org promoting governance…
@gleech — You funded ChinaTalk because it was the only publication covering China/AI with real practitioner access, in a language philanthropists couldn't evaluate. This project is the same logic for Taiwan.
Taiwan AI Governance Forum is the only Traditional Chinese-language AI governance community serving Taiwan's manufacturing sector — 7,224 active members, predominantly founders and CEOs,…
Ask Leo! He's giving 10k grants on here to multiple people.
Thank you for the tip @nicholas I just did, @LeoGao Nicholas deserves the $100 bounty for referral, if I'd get a microgrant.
Note: This grant was given as part of the experimental alignment microgrant program. Because this is a weird experimental grant program aiming to make lots of unusual bets, grantees are not allowed to use this grant as a credential when applying to any other grant, or job, or on their…
Note: This grant was given as part of the experimental alignment microgrant program. Because this is a weird experimental grant program aiming to make lots of unusual bets, grantees are not allowed to use this grant as a credential when applying to any other grant, or job, or on their…
2026 should be the year for AI Safety comms funding, but resources haven't caught up to the urgency. Petr is one of the leaders in this space and deserves the support to create the impact this moment demands.
@melgarcia Hey Garcia!
I didn't see any external contacts, so I am contacting you here. I agree with what you said about AI safety funding! While 2023 was the year of massive ChatGPT jailbreaks that I personally witnessed, 2026 was worse since people can now train their own models much more easily than in 2023 to commit cybercrime. With that being said, I was wondering if you had any…
Hope this gets funded soon! AI safety is really underrated among kids my age, and I know a lot of resources from my (admittedly) extensive research into grants. Let me know if you are interested!
Note: This grant was given as part of the experimental alignment microgrant program. Because this is a weird experimental grant program aiming to make lots of unusual bets, grantees are not allowed to use this grant as a credential when applying to any other grant, or job, or on their…
Note: This grant was given as part of the experimental alignment microgrant program. Because this is a weird experimental grant program aiming to make lots of unusual bets, grantees are not allowed to use this grant as a credential when applying to any other grant, or job, or on their…
Note: This grant was given as part of the experimental alignment microgrant program. Because this is a weird experimental grant program aiming to make lots of unusual bets, grantees are not allowed to use this grant as a credential when applying to any other grant, or job, or on their…
Heavy vouch for Petr! Communicators like Petr are rare to find and have a multiplier effect on the entire field. His first few YouTube videos on his new channel did extremely well- This is rare- it takes time to learn how to make good content, but with Petr's experience at Veritasium, and his undying passion and attitude for content- I wish that he is able to pursue content full time. Petr…
Quick flag: "The AI Safety Atlas" is already an established educational resource (the textbook/course used by thousands of students a year, and recommended by 80k career advisors). Reusing the exact name will likely cause confusion in the field. Would you consider a different name? "AI Safety Registry," "AI Safety Directory," or similar? The project itself…
@charbel-raphael Sure thing! Thanks for the note!
@Jesse-Richardson Following up with what’s developed since this was posted.
NIST’s National Cybersecurity Centre of Excellence published a concept paper in February naming exactly the failure mode this work measures. Non repudiation across multi hop agent delegation chains, flagged as an open problem, not a solved one. The three hop emergent authority finding in this dataset is the first…
[Final report]
I am closing this project as it has been subsumed into the ARIA Opportunity Seed on Mathematics for Safe AI: ARIA Opportunity Seeds – [SafePlanBench & Logically Constrained Reinforcement…
Applied to grantmaking.ai Launch Round · $5,000 requested
Theory of change, i.e. 'How does this work reduce existential risk from advanced AI?' AISafety.com is a field-building project that aims to multiply global AI safety efforts through a centralized hub that actively congregates and surfaces 'everything' useful AI safety. There are pages dedicated to events, training programs, self-study, mentors, and more (check out the site! And even use it),…
Get in contact with @OzzieGooen. Ill fund a continuation of what he has
@MarcusAbramovitch thanks for the pointer, I dug into Ozzie's work properly. I think we're solving different problems. His own vision doc (https://www.longtermwiki.com/wiki/E883) calls LongtermWiki "a strategic intelligence platform for AI safety prioritization" for funders and researchers asking "where should the next dollar or researcher-hour…
[Final report]
Thanks for your support of the Cooperative AI Research Fellowship! The program was a great success.
You can see the research outputs of the program here: https://www.cai-research-fellowship.com/posters/
You can see an abridged report here: https://canva.link/ci1hcmyjhzc0w6a
Any resources suggestions welcome.
This is a timely and ambitious project that addresses one of the most important governance challenges in the AI era.
As someone interested in digital health, I see significant value in this work. Making AI systems more transparent and easier to understand can help patients, healthcare professionals, and organizations make safer, more informed choices, while strengthening trust and…
Thanks so much for your kind words and support, @Beemetrics :)
You are very welcome @aashkapatel !
I’m currently building AI-ready clinical data infrastructure for underrepresented populations, I’ve published it here on Manifund and I would appreciate to have your feedback on that 😊
I think this is a good list of initial interviewees, and I think Steve Hsu's involvement is a strong signal for the success of this documentary. These folks screened their trailer at Manifest & held a q&a (which I sadly missed).
One worry I'd have is that a talking heads documentary like this (or the SB 1047 documentary) may have limited widespread appeal, esp compared to something…
I've been thinking about where this goes after the directory is solid and wanted to share it here.
The bigger goal is to make ATLAS the matching layer for the field, not just a place to browse:For hiring: a lab or recruiter posts a role and gets a ranked shortlist of people who actually fit, instead of searching. Same for new labs/startups that just want newcomers or fresh people in…
I've been thinking about where this goes after the directory is solid and wanted to share it here.
The bigger goal is to make ATLAS the matching layer for the field, not just a place to browse:For hiring: a lab or recruiter posts a role and gets a ranked shortlist of people who actually fit, instead of searching. Same for new labs/startups that just want newcomers or fresh people in…
Research update — June 2026
Since posting this project, I have continued the literature review and theoretical formulation. The project has narrowed into a more concrete technical question:
Can chain-of-thought reasoning be modelled as a learned local transition over hidden representation states, repeatedly applied across decoding time, and validated through token-level long-chain…
Significant research output since the original application.
The ARCS programme has published seven new papers on Zenodo this week covering findings that did not exist in the literature before this sprint.
The headline finding: emergent authority in three hop AI agent chains. In tested configurations, chain framing alone manufactured authorisation that no individual model in the chain would…
[Final report]
Here's the second progress report; in short, we generated the intended corpus and then a second bonus "victory-lap" corpus testing the Turntrout hypothesis. Here:…
Glad to answer any questions - today's decission to shut down Fable 5, shows that a general, non-agnostic, governance layer will be needed quite soon. It highlights the massive geopolitical and operational vulnerabilities of relying purely on centralized, platform-native AI governance.