Is not giving to X-risk or far future orgs for reasons of risk aversion selfish?

I have this idea which I haven’t fully fleshed out yet, but I’m look­ing to get some feed­back. To sim­plify this, I’ll em­body the idea in a sin­gle, hy­po­thet­i­cal Effec­tive Altru­ist called Alex. I’ll as­sume silly things like no in­fla­tion for sim­plic­ity. I also use ‘lives saved’ as a proxy for ‘good done’; al­though this is grossly over­sim­plified it doesn’t af­fect the ar­gu­ment.

Alex is earn­ing to give, and es­ti­mates that they will be able to give $1 mil­lion over their life­time. They have thought a lot about ex­is­ten­tial risk, and agree that re­duc­ing ex­is­ten­tial risk would be a good thing, and also agree that the prob­lem is at least par­tially tractable. Alex also ac­cepts things like the no­tion that fu­ture lives are equally as valuable as lives to­day. How­ever, Alex is some­what risk averse.

After care­ful mod­el­ling, Alex es­ti­mates that they could save a life for $4,000, and thus could save 250 lives over their own life­time. Alex also thinks that their $1 mil­lion might slightly re­duce the risk of some catas­trophic event, but it prob­a­bly won’t. On ex­pected value terms, they es­ti­mate that donat­ing to an X-risk or­gani­sa­tion is about ten times as good as donat­ing to a poverty char­ity (they es­ti­mate ‘sav­ing’ 2,500 lives on av­er­age).

How­ever, all things con­sid­ered, Alex still de­cides to donate to the poverty or­gani­sa­tion, be­cause they are risk averse, and the chances of them mak­ing a differ­ence by donat­ing to the X-risk or­gani­sa­tion are very low in­deed.

This seems to em­body the at­ti­tude of many EAs I know. How­ever, the ques­tion I’d like to pose is: is this self­ish?

It seems like some kind of moral nar­cis­sism to say that one would pre­fer to in­crease their chances of their per­sonal ac­tions mak­ing a differ­ence at the ex­pense of over­all wellbe­ing in ex­pec­ta­tion. If a world where ev­ery­one gave to X-risk meant a mean­ingful re­duc­tion in the prob­a­bil­ity of a catas­tro­phe, shouldn’t we all be work­ing to­wards that in­stead of try­ing to max­imise the chances that our per­sonal dol­lars make a differ­ence?

As I said, I’m still think­ing this through, and don’t mean to im­ply that any­one donat­ing to a poverty char­ity in­stead of an X-risk or­gani­sa­tion is self­ish. I’m very keen on crit­i­cism and feed­back here.

Things that would im­ply I’m wrong in­clude ex­is­ten­tial risk re­duc­tion not be­ing tractable or not be­ing good, some ar­gu­ment for risk aver­sion that I’m over­look­ing, an ar­gu­ment for dis­count­ing fu­ture life, or some­thing that doesn’t as­sume a hardline clas­si­cal he­do­nis­tic util­i­tar­ian take on ethics (or any­thing else I’ve over­looked).

For what it’s worth, my dona­tions to date have been over­whelm­ingly to poverty char­i­ties, so to date at least, I am Alex.