I recommend Model Meals as a partial substitute for cooking.
Most people can make decent progress to a healthy diet with just a little microwaving. E.g eat more nuts, whole fruit, sweet potatoes, and occasionally a can of oysters.
I recommend Model Meals as a partial substitute for cooking.
Most people can make decent progress to a healthy diet with just a little microwaving. E.g eat more nuts, whole fruit, sweet potatoes, and occasionally a can of oysters.
He says the “totally united by the same utility function” part is implausible:
he claims (quite implausibly I think) that all AGIs naturally coordinate to merge into a single system to defeat competition-based checks.
You’ve convinced me of your main points.
I recently noticed another complication for this analysis.
I’ve got a single hose AC that’s closer to the window than yours, with some gaps near the hose that let some air in from the outside. That means I’ve accidentally set it up to partly act as if it had two hoses. At one point the AC reported its intake temperature at 86F, while a distant part of the room was 80F, versus around 95F outside. [The AC is nowhere near as effective as these number suggest. Mostly the room has enough insulation to delay the effects of the outside temperature by hours.]
I asked myself for an example of a country whose initial pandemic response was unusually poor, settled on Brazil, and found that Brazil’s IQ was lower than I expected at 87. So that’s one data point that supports your hypothesis.
I suspect that cultural homogeneity is at least as important.
The beginning of this post seems fairly good.
I agree that an AGI would need lots of trial and error to develop a major new technology.
I’m unsure whether an AGI would need to be as slow as humans about that trial and error. If it needs secrecy, that might be a big constraint. If it gets human cooperation, I’d expect it to improve significantly on human R&D speed.
I also see a nontrivial chance that humans will develop Drexlerian nanotech before full AGI.
Your post gets stranger toward the end.
I don’t see much value in a careful engineering analysis of how an AGI might kill us. Most likely it would involve a complex set of strategies, including plenty of manipulation, with no one attack producing certain victory by itself, but with humanity being overwhelmed by the number of different kinds of attack. There’s enough uncertainty in that kind of fight that I don’t expect to get a consensus on who would win. The uncertainty ought to be scary enough that we shouldn’t need to prove who would win.
Iceland got tired of Europeans overlooking its talent, and moved to a more conspicuous location.
Strong upvote for tackling an important problem. I’ve tried to write something along these lines, and made no progress.
But I still see lots of room for improvement.
I’d like to see some more sophisticated versions of the Turing tests: use judges that have a decent track record on Turing tests, and have them take longer than 2 hours.
I don’t think the Nano AGI test should rely on statistical significance—that says more about the sample size than about the effect size.
I agree with some of your complaints here. But Eliezer has more of a track record than you indicate. E.g. he made one attempt that I know of to time the stock market, buying on March 23, 2020 - the day on which it made its low for 2020.
Global warming accelerates due to waste heat. See Freitas, Global Hypsithermal Limit for some relevant estimates.
Paul Christiano made some estimates here.
Soil depletion seems to have caused enough problems in the past that I don’t find the lack of recent civilization collapse very reassuring.
I looked into it and posted about it here. It’s not currently a crisis. I have some medium-sized concerns that it’s being poorly managed.
My intuition told me that Jürgen Schmidhuber likely did, and Google led me to his Formal Theory of Creativity and Fun and Intrinsic Motivation. I have not read it.
I don’t expect to find strong evidence on this topic anytime soon, so I’m making do with what’s available. I think I’ve been influenced by a fair amount of poorly legible evidence.
Cortisol and HRV data would likely be valuable. I predict that they would show that the average Amish person is mildly less stressed than the average American.
The Amish seem to approve of outsiders converting to be Amish, but they are definitely not making it easy. It’s likely somewhat hard to get started, because more tourists want to visit Amish communities than those communities are willing to interact with. You’d need to learn an obscure dialect of German in order to be accepted. Then learn lots of rules about which technologies can be used when. Many people are mildly addicted to something that the Amish prohibit or heavily restrict (television, electricity).
Something like half of new converts eventually drop out. That’s important evidence that establishes that Amish communities can’t be a lot better than other cultures. But I think it’s consistent with Amish culture being slightly better on average.
I’m only making a rather weak claim that Amish communities are a decent place to live. My stronger claim is that, based on what many Americans claim to want, there’s something weird about how few give any thought to converting. In particular, I see a discrepancy between how much people say they value equality, versus how much they seek it out.
The evidence on sexual abuse seems to be entirely anecdotal, and not quantified in any meaningful sense. I included it to avoid implying that the Amish are consistently minimizing crime.
Adolescents during Rumspringa are certainly subject to social pressure, but not obviously more so than in other cultures. They appear to have less parental control than those with tiger moms.
Here’s a survey of Amish women that tends to confirm the mostly good outcomes: Health status, health conditions, and health behaviors among Amish women (ungated copy). For children, the high retention rates seem like medium-quality evidence.
I wonder how much their satisfaction depends on the hope of eternity in heaven.
I agree with most of this advice. I probably couldn’t do any better than that. But it seems unlikely to be the best that’s possible. Dementia can probably be stopped in early stages if there’s a way to persuade the patient to make larger lifestyle changes. It’s frustrating that such persuasion is unusually hard.
On oil production: my guess is that they’d increase oil production if futures in the 2025 − 2030 range were $90+. Those futures are currently around $65-$70.
Drexler wrote his QNR paper in part to address this issue. I’m trying to write a blog post about QNR.