They settled on Condorcet, a method almost unheard of outside of electoral/political theory, and your criticism is that they didn't make use of political theory?!
It's training data. There's 17 suicidal and 17 non-suicidal scans, for a total of 34 scans. They trained 34 models, leaving one scan out each time. Of those 34 models, 31 correctly predicted the left-out scan.
IANAStatistician, but this seems like a trash result.
Letting people make mistakes is important. Letting people be altruistic is important.
More specifically: I'm 0% worried about sending money overseas, since I expect that the few people who do so that will be making the world a better place on net.
Nothing is free - someone is paying for it. When you give people specific things instead of money, you're taking choices away from them, insisting that you know what they need better than they do.
If you give people money, they can decide what they need most and go do that. Maybe I don't need housing because I can live with relatives and use my money for education. If you insist on free housing instead, you take that option away from me.
"IQ" is a scalar semi-objectice well-defined metric that predicts some things about the much more complicated and less-well-defined concept of problem solving.
That post is a remarkable example of victim blaming. "Do you know how hard it is for ME now that you've gone and gotten people violently angry at you? People want to punch you in the face for something they think you said, and that's your fault!"
Yes, so the rule is slightly misstated. Assuming that your messages are all under n bits and you always use an n-length pad, you will reuse the same pad approximately once every 2^n messages. Not coincidentally, 2^n is also the number of possible one-time pads there are, so your adversary gains no information from this fact.
I routinely pay for things with ethereum - and if the target only accepts btc, then I can just shapeshift.
The "Halting Problem" is a red herring here - eth transactions are technically not Turing complete since they automatically halt when they run out of the finite supply of ether attached to the transaction. Saying that ethereum isn't feasible because of the halting problem is like saying cars aren't feasible because an object in motion stays in motion - there are countless practical ways to manage it, and if nothing else stops you, in the end you'll definitely run out of gas.
>
Moreover, to add further context, the whole basis of Holden's effective altruism work has been around the idea that philanthropic dollars ought to be focused on charities with extremely rigorous proof behind how much they improve people's lives per dollar donated, and how much they need the money.
This isn't quite true; SCI, a charity that treats parasitic disease in the third world, is the subject of massive uncertainty and conflicting reports of effectiveness. It might turn out that it has very little impact at all. But it's still a recommended EA charity because it looks like there's a decent chance they're doing a ton of good. GW has written extensively about this.
> The Open Philanthropy Project typically recommends grants to the Open Philanthropy Project fund, a donor advised fund at the Silicon Valley Community Foundation. Support for the Open Philanthropy Project fund comes primarily from Good Ventures, though other donors have contributed as well. In some cases, the Open Philanthropy Project makes grant recommendations directly to Good Ventures.
It's basically all GV money - that is to say, Dustin and Cari's money.
It would be really useful if there were some historical example of a sudden increase in intelligence leading to a new class of agents taking over the Earth and determining it's future.
Or maybe if there were any evidence that AI capabilities are increasing, like becoming dominant at Chess or Go, driverless cars, or a slew of recent papers on transfer learning.
Maybe if one of the co-authors of the leading AI textbook, Stuart Russell, voiced concerns, we could count that as evidence.
But you're right, it's better to wait until we know for a fact that someone's built an agent smart enough to end civilization before we commit any resources to the problem.