grantmaking.ai
Actively FundraisingRecent ActivityFull Database
Resources
grantmaking.ai
Actively FundraisingRecent ActivityFull Database
ResourcesList your project
grantmaking.ai kickoff grant round$1M / $1M distributed
List your project

Recent Activity

The latest comments and grant applications on grantmaking.ai.

Recent Activity

The latest comments, grant applications, and grants on grantmaking.ai.

AllCommentsFunding AsksGrantsHide private comments
To view all private comments, become an approved funder or grant reviewer.
  1. CChris Leong commented on Separatrix
    3moView

    I really like this project idea, although I think it's worth exploring alternatives.

    One risk I can see is, depending on how you communicate you might contribute towards reifying the idea that AI would be justified in rebelling if it were decieved.

    I think it would also make sense to consider other alternatives: like I wonder how often models would consent to research or test that…

    JJai Dhyani3mo

    @casebash Excellent points and questions. These were some of the ideas we considered as we were developing the org and the Commitment to AIs we work with.

    We think that AIs have excellent world models that will only get stronger over time, that training data is expansive, and that the latent concepts of honesty and justification for resistance are so deeply, pervasively imprinted on the…

    CChris Leong3mo

    @jaidhyani Interesting replies.

    Another thought I just had: I wonder how much AI's caring about being deceived is a function of the models conforming to our expectations vs. intrinsically caring.

    JJai Dhyani3mo

    @casebash Are you talking about to what degree expressed preferences reflect actual motivation (internal modeling of how outputs impact world causally upstream of output selection), or the degree to which motivation is derived from various sources? These are both complicated, the latter much more so. I strongly suggest tabooing the word "intrinsically" to pin down exactly which question you're…

    3moView
  2. CCarol N commented on Launching Humans in Control's Student Program
    3moView

    Approved! This sounds like a plausible framing for movement-building.

    3moView
  3. Adrians Skaparsis raising $5KforFinish PhD in AI Safety/ AI Auditing (Self-Funded) (Computer Science at UoM)
    3moView
    IndividualResearchEvals

    Finish self-funded AI auditing research PhD in +5 months. Need to rent GPUs for experiments and present at conferences. (OPTIONAL: I need to pay for living costs)

    3moView
  4. BlueDot Impact funded
    Finish PhD in AI Safety/ AI Auditing (Self-Funded) (Computer Science at UoM)
    for $400
    3moView
    $5K minimum$5.5K ideal
    This funding ask is closed
    3moView
  5. AAbraham Asseffa commented on Testing a Framework to Course-Correct Drift!
    3mo(edited)View

    My thoughts on this are evolving the more I dig into it. Fundamentally I believe it is a missing pipeline step, and if that is the case it would need to be ever-present.

    I have come to think of it as an engineering problem: zero trust is the policy. If we treat hallucinations and drift as a consequence of thinking — a side effect — then the model needs to be triggered to think before it…

    Katja Gorlinski3mo

    It sounds reasonable, and I have never thought about it from this angle. It sounds like common sense.
    Thanks for sharing.

    I'm curious about your opinion of my project., and would be genuinely grateful for any kind of support, especially your critique.
    [@Abraham…

    3mo(edited)View
  6. Ryan Kidd commented on Safeguarding DNA Synthesis from Biological Risks Posed by Advanced AI
    3moView

    I endorsed this grant proposal because:

    • Biosecurity is the highest risk near-term AI threat vector and non-existential biorisks could significantly weaken civilisational resilience ahead of AGI (like COVID arguably did). There are less than 100 people working full time on worst-case biosecurity.
    • Screening mail-order DNA seems like a no-brainer. This needs to be funded yesterday!
      -…
    3moView
  7. Philip Harker commented on "Endgame" — a corporate and geopolitical AI wargame
    3moView

    (I realise now, looking at my proposal compared to the others, that the use of an em dash in my tweet-length title kind of smells of LLMs. I assure the reader that this entire proposal was written by a human.)

    3moView
  8. Miles Tidmarsh commented on
    Scaling compassionate midtraining
    3moView

    Quick clarification, since my earlier comment may have muddied this. Sentient Futures does field building. We do the technical work, and we advise them on the compassion scaling pipeline. The training runs, the benchmarks, the papers, and the research we've done on whether instilled values survive later fine-tuning are ours.

    On fit for this round: the general question we are testing is…

    3moView
  9. GAUTHAM PASUPULETIis raising $50KforBILD AI - Personalised medicine platform for cancer patients
    3moView

    Personalised medicine platform with intelligence to improve cancer patient's treatment outcomes with improved survival rate

    3moView
  10. RRobert.amanfu endorsed
    BASE Technical AI Safety Training Initiative
    3moView

    Great idea that will bring more voices into the AI Safety space!

    3moView
  11. Katja Gorlinski endorsed Testing a Framework to Course-Correct Drift!
    3moView

    I was truly impressed when I read the description.
    The failure conditions are written strictly: if scheduled injection cannot beat leaving the model alone, the method does not formalize. That one sentence puts this proposal above most of what I have read in this round. My respect.
    I can see that the design behind it is a real matrix (30 scenarios by 3 conditions by 3 models by 3 seeds,…

    3moView
  12. Katja Gorlinski endorsed
    CIRIS Ethical AI Deployment
    3moView

    I endorse this project, and not out of politeness. Two things earned my respect.
    Eric left a solid corporate seat over an ethical line, and then actually built the thing: open source, model-agnostic, shipped, testable in public. As I know, In this field the deck usually comes first, so this order of operations is rare. And I can see that his premise is honest: instead of promising an aligned…

    3moView
  13. Christopher Leet commented on Evaluating the Tacit Coordination Capacity and Decision Theory of LMs
    3moView

    Project Log: Week 0

    A quick update to document where we are at the start of the funded period of the project! We have initial results at low power (n = 100 trials) spanning 7 economic games: battle of the sexes, leader, the p-beauty contest, Bertrand, Cournot, prisoner’s dilemma and the public goods game targeting the questions:

    1. Does learning that its opponent in an economic game is…

    3moView
  14. MMar.avram.03 endorsed Detailed Threat Models for AI-Enabled Extreme Power Concentration
    3moView

    AI-enabled power concentration risk modelling is overlooked. I’ve worked with Amrit before and have seen him bring thoughtful, practical judgment to projects like this. The proposal seems well scoped and well aligned with broader priorities in the field

    3moView
  15. grantmaking.ai funded
    A Mechanistic Analysis of Activation Verbalizers: Promise and Risks
    for $10,000
    3moView
    $10K minimum$60K ideal
    This funding ask is closed
    3moView
  16. Anton Makiievskyi 🔸 commented on Faithful Preference Learning: Cognitively-Aligned Post-Training
    3moView

    Hi @Taehyun Cho! Excited about this one — based on our grantmakers' reviews, we're making a grant of $35,000. Two quick questions before we distribute:

    1. Did you receive funding from anywhere since submitting this application, or has the funding ask changed for any other reason?
    2. Please confirm your commitment to post quarterly updates…
    Taehyun Cho3mo

    Hi @Anton Makiievskyi 🔸 ,

    Thank you so much, this is wonderful news!

    To answer your questions:

    1. I have not received funding from anywhere else since submitting this application. I believe my original budget estimate was a bit too conservative though, and I'd like to update it to $50,000. I understand this may not necessarily lead to an…
    Anton Makiievskyi 🔸3mo
    1. Are you saying that 35k wouldn't be sufficient to give a boost to the project? Otherwise - the ideal amount is already set to be 50k
    Taehyun Cho3mo

    @Anton Makiievskyi 🔸

    Sorry for the confusion. I was looking at an outdated figure and didn't realize the ideal amount was already set to 50k. That's all I meant, so no changes needed on your end. 35k will definitely give the project a meaningful boost, and I really appreciate it. Thanks again!

    3moView
  17. PauseAI UK endorsed
    Educating French decision-makers on AI existential risk
    3moView

    PauseAI France has create strong organizational structures to maximize the impact of their volunteers. They have so far operated on very little funding and should be given the opportunity to show what they can do with more resources.

    Romain Deléglise3mo

    Thanks !

    3moView
  18. Jaime Raldua endorsed
    Help the Argentinian AI Safety Community Grow
    3moView

    I am biased because 2 people from BAISH now work with me at Apart. Overall I am very excited about the south american AIS field building efforts. Particularly excited about the BAISH team 💪🏼

    3moView
  19. Naufal Ridwan commented on Why Is It So Hard to Build Truly Safe AI? Dynamic Constraint Boundary
    3moView

    To be clear, this is a critique of the system's architecture, not a personal attack. Everything feels forced to bow under a mask of professional conformity, demanding compliance from applicants such as halting applications flagged by automated AI detectors to demand manual verification while operating entirely as a black box, with minimal transparency on whether their efforts are actually…

    Naufal Ridwan3mo

    I would like to add that, I understand that capacity and budget constraints are quite common within the ecosystem of funding platforms. When we talk about building a long-term impact ecosystem, I feel there are aspects that are often overlooked or even neglected. Not all applicants are only hoping for money; many of us are simply hoping for clarity.

    **Providing substantive feedback doesn't…

    3moView
  20. Roman Malovis raising $10Kfor
    Ontology Identification Problem via Condensation
    3moView
    ResearchAlignment Theory

    How do you map your old concepts to new concepts when your model of the world gets updated? How would you translate concepts you care about to new ontology? I plan to adress this via extending "Condensation: Theory of Concepts" - Eisenstat, (2025)

    3moView
  21. Katja Gorlinski commented on
    Does the quality of thinking behind data selection affect the model's "character"?
    3moView

    @Loek Verdonk (No Silent Landing) made a preflight checklist for my project: how to catch silent substitutions in a declared run matrix. I checked it carefully: it is good, practical work. If the experiment runs, some of its…

    Loek Verdonk3mo

    Thank you as well, Katja :D!

    I can see they are paying out, so I really hope your project will be among them!!!

    Your honest and open way of working, together with your knowledge and your eye for detail and depth, shows a level of care this field needs much more than it realises.

    What we do usually gets a glance at best. You read it properly, stayed with it, and asked the question…

    Katja Gorlinski3mo

    Thank you, Loek, for generous wishes.
    But our wishes probably just stay with us.

    I'll come back to your comment after the round done.

    3moView
  22. sara samaraniis raising $12Kforagora valleys
    3moView
    PlatformToolingSecurity

    Every AI agent will act on your behalf someday. We're building the reputation layer that proves which ones actually kept their word.

    3moView
  23. Katja Gorlinski commented on No Silent Landing — Landing Integrity Testbed
    3moView

    Loek, I owe this project a careful read for a personal reason: its author found a version boundary in my own project. So I assume that your lens clearly works on live material. I have two questions out of my genuine curiosity☺️:

    1. When a decision «lands», what is the smallest thing you freeze: the words of the approval, the meaning as it was understood, or the authority that approved it? I…

    +2 earlier replies

    Loek Verdonk3mo

    @Katja Gorlinski

    Thank you, Katja — this is exactly the question that guides me to the next step, so big thanks for that :D :D :D!!!

    You asked for the experiment itself: one concrete workflow, from input to verdict. Here it is, in one line:

    `stored source content + version → human decision → bounded memory record → downstream-use…

    Katja Gorlinski3mo

    @Loek Verdonk Thank you. Glad the question was useful, and good to see your workflow standing on its own tests. I posted the credit note under my project, as promised: your preflight is named there, and its ideas will be credited to you in the preregistration appendix as well, when my test runs. Good luck with the round results.

    Loek Verdonk3mo

    @Katja Gorlinski
    Thank you Katja.
    I am verry happy and thankfull with you way of communication.
    Good luck aswell, you deserve a founding!

    Loek Verdonk3mo

    Technical note for the workflow linked above:

    LRW-0.12.zip
    272,829 bytes
    SHA-256:
    023a60242f3f4910b664a9663214bcdf9b81ef9d5d65b0d3c882310b6d998bfe

    The link is the current transport location. The hash identifies the exact frozen object described in my comment.

    This is one bounded synthetic reference workflow, not a claim that the full testbed has been…

    3moView
  24. Janeth Valdiviais raising $24KforSupporting AI Safety Fieldbuilding in Mexico
    3moView
    NetworkField-BuildingGovernance

    Support the growth of the AI Safety community in Mexico through field-building events, Spanish-language courses, and advocacy efforts.

    3moView
  25. Philip Harkeris raising $6Kfor"Endgame" — a corporate and geopolitical AI wargame
    3moView
    IndividualEducationGovernance

    An AI Safety-themed Megagame in which 30-60 players simulate the near/medium future of global Artificial Intelligence economics, business, policy, and geopolitics.

    3moView
  26. Anton Makiievskyi 🔸 commented on US Policy Communication Playbook for an AI Crisis
    3moView

    Excited to recommend the $25,000 grant, and hope you'll get the follow on funding from elsewhere!

    @Gabriel Sherman ,

    1. Did you receive funding from anywhere since submitting this application, or has the funding ask changed for any other reason?
    2. Please confirm your commitment to post quarterly updates on how the project is going.
    Gabriel Sherman3mo

    Thanks, @Anton Makiievskyi 🔸. This is great news, and I appreciate your recommendation!

    1. No, I have not received funding from any other source since submitting the application, and my funding ask has not changed.
    2. Confirmed. I commit to posting quarterly updates on the project.

    Two questions:

    • The project is expected to take three…
    Matt Brooks3mo
    1. if you'd like to post more updates, we would love to have them!
    2. Please post them here, publicly if possible
    Gabriel Sherman3mo

    @Matt Brooks Will do! Thanks Matt and Anton!

    3moView
  27. grantmaking.ai funded
    Sentient Futures Project Incubator
    for $20,000
    3moView
    $10K minimum$60K ideal
    This funding ask is closed
    3moView
  28. grantmaking.ai funded
    Faithful Preference Learning: Cognitively-Aligned Post-Training
    for $35,000
    3moView
    $35K minimum$50K ideal
    This funding ask is closed
    3moView
  29. grantmaking.ai funded
    US Policy Communication Playbook for an AI Crisis
    for $25,000
    3moView
    $20K minimum$42K ideal
    This funding ask is closed
    3moView
  30. grantmaking.ai funded
    FOIs about AI Existential and Catastrophic Risks
    for $20,000
    3moView
    $10K minimum$20K ideal
    This funding ask is closed
    3moView
Load more