While I see this as a meaningful threat, I think that it’s something that pretty smoothly follows from the development of mind-reading at any point in time, and so a five or twenty year delay is not particularly relevant to it.
KellerScholl
Karma: 622
I’m curious what sort of work you expect to be useful here. “Overthrow non-democratic regimes in advance” has other drawbacks, and doesn’t seem to have the characteristics you’re talking about.
“I make myself coffee. My thoughts are mostly empty, but I briefly recall what I want to say to the candidate I’m getting lunch with tomorrow. In the background, Codex is prompted to think about info I might want for the lunch chat, and decides that updating the synthetic data scaling plot before showing it off would be prudent; Codex spins off a subagent.
I’m not saying words really loudly in my head while getting coffee. I just read the plots as I normally do, and make coffee as I normally do. It feels like magic.”
I’m quoting from the post I reference at the top. A deliberate writing decision I made was not to question the feasibility of the technology: if it’s not feasible then it’s not worth putting effort into it, but this is not my area of expertise and “this novel technology might not work” is not particularly interesting.
“How much alcohol has this person consumed recently” and “are you loyal to the Glorious Leader” are different questions, and truth machines for them will have different effects.
I genuinely believe that it’s a good thing for the majority religion of the country to have at least one representative on the Supreme Court who understands it in the way that it can be difficult for a non-native. I think it’s particularly bad for legitimacy to have an entirely Catholic and Jewish court. I agree the statement is structured as a joke, but I believe there should ideally be at least one man, at least one woman, at least one Protestant, and I know that this one is more of a stretch for lifecycle and class reasons but I’d like to have at least one renter (one third of American households), at least not a geographic mono-culture, etc.
https://x.com/JacobHHilton/status/1794090561294467074
He has also explicitly told people not to expect candor from him on this issue until the situation changes. That the binding is no longer part of a contract, as opposed to implicit threat, seems of little relevance.
I am very curious how you think about this post in retrospect: parts of it seem clearly falsified. I completely understand if you currently feel bound by a non-disparagement clause and expect it to be a few weeks before that can be confirmed to no longer apply.
Three years later, I mostly stand by this: war with China has not happened, but estimates of likelihoods have risen, and it not driven by food concerns in the slightest.
How do you expect journalism to work? The author is trying to contribute one specific story, in detail. Readers have other experiences to compare and draw from. If this was an academic piece, I might be more sympathetic.
Correction: the Youtube link should point to https://www.youtube.com/watch?v=ZpwSNiLV-nw, not the current location (a previous video of yours).
I wish I knew! Nobody has yet explained it to me, nor do I have any theories I am particularly confident in.
This is deeply unconvincing. We didn’t have a great power war in the 60s or the 70s because that would have meant nuclear war. High-level US government officials in internal documents describe Russia as an existential threat. Russian government documents, as I understand it, reflect terror of American willingness to use nukes. We haven’t had a war between the US and China yet, but estimates of that holding true over the next five years are less confident than I’d like.
“Most wars have ultimately been fought over land because land determines food production and food production was a matter of life and death.”
It seems like you’re explaining the actions of kings with the preferences of peasants (and I am very unconvinced that a victorious war was better for the average peasant than peace), and I don’t see that as particularly persuasive.
Priors are relative to how much evidence can be shared. There may not be agreement in a single conversation, but they should expect movement towards a common belief, though there are degenerate counter-cases. For example, perhaps both parties share a base rate and have different pieces of information that push in the same direction.
I think that the reason I don’t see a lot of arguments against anti-vaxxers is that I don’t know that I know of any. I think the reason that I see anti-vaxxers derided more often than average is flat-earthers are parsed as harmless and anti-vaxxers are parsed as doing harm. I think I’m not quite following what you’re saying.
There’s not a hard cutoff between 2005, when Ioannidis publishes, and the present, but I’ve worked on multiple systematic reviews, going over thousands of papers, and there’s a visible improvement in quality over time, and that seemed like a reasonable date for “replication crisis attention is high.”
This seems like an excessively general question to me. Yes, because we’ve gotten richer over time. No, because there is still suffering. Can you drill down into specifics?
Instead, we should tax the difference between what you earned and what anyone could have made by just putting the same amount of money in a savings account. That is, we should tax the stuff that is actually income, i.e. when you are actually doing work, taking risks, or exploiting connections.
I think that “savings account” is an underdefined term here, which I think causes serious problems. “doing work” and “taking risks” seem like income, and I see the argument for taxing them accordingly. Does “taking risks” mean “US Treasury Bonds” (which have a risk of default)? “broad market indices”? “employee stock options”? I think I would say that the market overall doesn’t count as income in a meaningful sense, but that is very debatable.
I think an advantage of carry-forward is that someone can’t get paid their marginal income tax rate for losing money. Marginal income tax at the moment caps at 50.3% in California (State+Federal), while long-term capital gains tax is 20%. There are a lot of accounting shenanigans that will make sense at a 45% rebate that would not at 5% rebate (approximating 10% annual returns and loss at halfway through the year, since carry-forward takes at least a year to come into affect assuming it is profits in the subsequent year that are at play, but even 15% vs 45% seems like more than enough room).
In advance of other comments:
1. Declining marginal utility of specific goods and non-uniform initial distributions of goods over people (this one matters).
2. There is a finite length of the production chain that one person can accomplish if something would take longer than an entire human lifetime to produce. Suppose I luck into a massive amount of unobtanium, with perfect property rights. To some extent this is 1, but I might also desire goods that I could not produce in an individual life, and by trade acquire them (this one isn’t that critical).
3. Building on 2, some goods are only possible at the end of extremely expensive or complex production chains: effectively high startup costs imply that not everyone should produce a good. For example, “an airplane” is something that a small fraction of is not nearly as valuable as the whole entity, and there are small enough fractions that are worthless as airplanes. (you might consider this solved by star-trek transporters, which deal with a similar problem and you thus consider the class of problems solved. That said, I consider this a key source of gains from trade)
4. There’s a more pedantic point that as an economist I feel obligated to make: Increasing marginal utility of two goods over a relevant range implies gains from trade (this one doesn’t matter for most understandings of trade)
After reading other comments:
Jsevillamol has a good point about insurance that I missed. Villam has a thoughtful point about R&D that I think is covered by the Matrix but arguably counts. Measure’s point about shared equipment relates to time-sharing and startup costs, though framed differently.
This seems very questionable: “does X matter?” is comparable to “is X vs not-X worth the cost of investigation?” If I’m constrained by resource limitations, and trying to acquire as much knowledge as I can given that, the ability to dismiss some answers as unimportant is critical.
Apologies for the delay: this is a long and thoughtful response I sadly don’t have the capacity to do justice to right now. Your iff seems to be central to your argument, but I am not convinced by either component. Firstly, for human + AI < AI to hold, we need AI to be less competent than the most competent humans, in any respect. It’s not obvious that AI alignment is at the right level of difficulty! Secondly, for human + AI + mindreading > AI to hold, it seems that AI must add a great deal of productivity. Why should I expect my subconscious thoughts and reflexes to be more like a 300% boost (for the typical employable researcher) than a 30% boost?