14Joined Aug 2019


Thanks so much for writing up this detailed post! Has anyone done the PointsBet promo? It's in a different format, so I can't directly use OddsJam. I'd really appreciate some guidance on how to make reasonable bets there.

I'm not sure if I understand this:

"There are two favorable variations of the risk-free first bet: The risk-free first bet, win or lose, (as offered on Caesars and BetFred) in which case your first bet should actually be made on short odds to reduce variance and tax liability. Or a risk-free first bet where the site credit returns the stake (as offered on Barstool and Fanduel) in which case your free bet should be made on short odds for the same reasons."

Following the paragraphs above, shouldn't one of these be long odds?

1. I think it is much more likely that different states should be counted according to their measure, not their existence. Denying this has the issues with preferred bases that you mentioned--since |+> = 1/sqrt(2) (|0> + |1>), it's unclear whether we should count one observer in state |+> or two observers, one in state |0> and one in state |1>--whereas accepting it will at least be consistent on how much moral weight the system has (weight <+|+> = 1 in the first case, weight 1/sqrt(2)(<0| + <1|) 1/sqrt(2)(|0> + |1>)  = 1 in the second case). (Also, this issue is even worse in infinite-dimensional Hilbert spaces--a position eigenstate is a superposition of infinitely many momentum eigenstates. If we counted each state in a superposition separately, the most important question in the world for suffering would be whether space is discrete.)

2. This isn't an issue that's unique to quantum computers in all interpretations of quantum mechanics. In a theory where wavefunction collapse is a physical process, then indeed quantum computers would be special; but in many worlds, everything is in a superposition anyway, and quantum computers are only special because they avoid decoherence at larger scales. I personally think something like many worlds is far more likely to be true than a theory that involves true nondeterminism, although it's not a closed question.

Thanks for this post! I was wondering if you had looked into ACT Grants (https://actgrants.in). When I looked into this on my own a few days ago, they seemed promising for several reasons:
- They're focusing on oxygen concentrators right now, which suggests that they are prioritizing well
- Their impact report is pretty impressive; most charities in this space don't seem to have impact reports at all
- Because of the above two, I am more confident that they would be able to use my donations well if the funding goals for concentrators are reached
- They're partnered with Swasth, which has a longer track record
I'd be curious to hear about it if anyone else has looked into this!

I agree with the first example--I want to help people effectively because I want to help myself effectively, and I have a deep conviction that others' pain is exactly as important as my own.