There are several reasons, but I agree that the most impactful research is most likely be mostly elsewhere. I don't think that in the scale we are looking for right now there is a lot of direct impact, and I view this as a learning opportunity - if this succeeds then it should definitely be scaled up globally or to other local groups, depends on the way it has succeeded.

Some reasons -
we are in better position to find and help Israeli people who are only partially aligned with EA, or perhaps never heard of it, to gradually come closer to EA or a promising cause area.
We want to make better connections with senior academic researchers and to generally have a good map of the relevant research spaces in Israel; we think that a good way to affect them is to nudge students a bit in the direction of more impactful research.
We might have more opportunities from donors who are looking to fund Israel-specific projects.

Good point. I mostly addressed the initial runs for this project, for which I actually don't think it's likely that either of these would be the most valuable. On the long term, from the experience of Effective Thesis and other conversations, I think that the most impact would come from connecting people to EA causes and helping them to start a career in those fields.

Yea, could be. I think that would depend on the individual case, but that we should probably prioritize more certain research projects.

Especially in your pilot round, it's probably better to err on the smaller side. I don't think missing promising opportunities is that important in Round 1, because I would imagine the experimental value (e.g. is this a good idea or not) would be the most important

Yea, that's a good consideration. My hesitation with only five is that then it might too easily be seen as either a success or a failure just by luck.

Is this because you think the supervision is useful enough to be mandated or because you think that the project itself should pass through experts? (the second can be done via the research proposal process)

For small projects like these, I think supervision and coordination / connection with an expert is likely the most valuable part in trying to get someone more involved in EA research usefully.

Agree. Edited to include that

In terms of small project benefits, having a grant on a student CV is very helpful when applying for later positions/fellowships.

I second that.
I think that to find really good applicants, you need to reach many people, and the university won't do this for you by default (and 800 people of different levels of engagement probably isn't enough)

I have concerns that this will make it more difficult for people to submit proposal if they are not sure

Yea, that makes sense

yea, it's important to hash out the details well, but I dont think it's crucial for understanding if we plan on going on with the project

Interesting. I think that is less important in the grant aspect, but it is relevant in the mentoring and direction aspect.

I agree. I'll start working on these connections once we get the go ahead with this project

We can find someone to fundraise :)

I'm certainly willing to evaluate technical AI alignment proposals, if such are received.

Thank you both! I really appreciate it.

Added previously that I think we should give the grant right at the get-go. (I find it difficult to really track people who might cheat, and I FIND it less likely anyway)

I didn't mean that people will cheat. I'm just assuming that some (most?) people won't finish their project, for whatever reasons.

I'm not sure how the process of grantmaking works. Does it count as income? Is it tax-deductible? can we give grants to people employed? Could people sue us for something silly?

I can imagine that Universities or professors object to overtly giving students salary to work on things other than their main research project. In my experience (from Australia), small 'salary' grants for students who already have a primary stipend are usually considered 'top-ups' to encourage them, but don't require them to work on anything specific beyond their main research project.

Yes, this would be a problem for students who are already receiving grants that cover all of their time - I don't think that would describe most applicants, though. Undergraduates who aren't funded, certainly.

Here I ment only people who are familiar with EA and consider themselves members of the community

regarding your question, I think that would generally be easier if someone is picking a problem from the broad EA research agenda, but if someone brings evidence that something we are not focused on might be important, that could be good to fund someone to work on it (mostly from a VOI perspective)