If you have any questions for me or just want to talk, feel free to reach out by sending a private message on this site or by sending an e-mail to egeerdil96@gmail.com.
You can also find me on Metaculus at https://www.metaculus.com/accounts/profile/116023/.
This may be the most underappreciated post on LessWrong that I’ve come across thus far.
I often hear that the Federal Reserve can’t hike interest rates too much (perhaps not enough to control inflation) because then the federal government would be unable to service its debts. Is this actually true?
The Fisher relation says such a hike has to either raise real interest rates or raise expected inflation. If you believe the interest rate hikes would go more into real interest rates than inflation then worrying about this is justified in principle, but in general there’s no good explanation of why the central bank raising nominal interest rates would raise real interest rates for a long time, say more than the few years it might take for prices and wages to adjust.
If the hike goes into expected inflation instead, then it’s true that the US government has to pay more nominal interest to roll over short-term debts, but they also collect more taxes because of inflation as the tax base becomes larger in nominal terms. Inflation is actually a fiscal benefit to the government in this case, as it ends up devaluing the long-term debt that the government owes.
The US government is sufficiently creditworthy that a few years of higher real interest rates would most likely not leave it in any danger of being unable to service its debts, but I admit there is some small risk every time real interest rates go up that markets end up losing confidence in the US government and refuse to roll over the trillions of dollars of Treasury bills that would need to be rolled over to prevent a debt default, either explicitly or implicitly through an inflating away of nominal debts.
Link to explanation of how/why this would work?
Market monetarists are one group that has been vocal in recommending this kind of policy, specifically with NGDP futures. From the wiki page:
Market monetarists advocate that the central bank clearly express an NGDP target (such as 5–6 percent annual NGDP growth in ordinary times) and for the central bank to use its policy tools to adjust NGDP until NGDP futures markets predict that the target will be achieved.
Alternatively, the central bank could let markets do the work. The bank would offer to buy and sell NGDP futures contracts at a price that would change at the same rate as the NGDP target. Investors would initiate trades as long as they saw profit opportunities from NGDP growth above (or below) the target. The money supply and interest rates would adjust to the point where markets expected NGDP to reach the target. These “open market operation”s (OMOs) would automatically tighten or loosen the money supply and raise or lower interest rates. The bank’s role is purely passive, buying or selling the contracts. This would partially or completely replace other bank’s use of interest rates, quantitative easing, etc., to intervene in the economy.
There have been proposals to do this, but in practice I think getting rid of central bankers and replacing them entirely with programs would not work so well. The problem is you often need humans to react to unexpected situations—a financial crisis, for instance.
I think routine implementation of monetary policy could actually be largely automated if only there were CPI futures or NGDP futures available in sufficient liquidity. The central bank could subsidize the creation of these markets if they thought it to be necessary, but central banks are bureaucratic organizations and so generally don’t want to take actions that would limit their discretion.
For similar reasons, central banks often have a difficult time being seen as credible by market participants when they attempt to commit to some policy on a medium to long time horizon, as they really have no formal process which would bind them to a promise they had given years ago.
In theory, of course, you can also privatize money if you think market incentives and competition would result in more efficient money management. My guess is there would be a strong pressure for private monies to aggregate together until a small number of them accounted for most of transaction volume and money balances, and then the institutions managing them (which could e.g. be a consortium of banks) would still be faced with the same monetary policy problems that central banks are faced with today. Still, policy could improve if there were some mechanism for people to dump poorly managed monies and transact in better managed ones.
That’s right :)
There’s a type of problem that is recognizably a “1/e problem”, in the sense that you expect the answer will somehow involve 1/e, but I haven’t been able to make this intuition precise. What properties about a problem make it likely that 1/e shows up?
Some of the above problems involve use of the inclusion-exclusion principle, which is a typical way 1/e can show up in these problems, but e.g. the secretary problem does not.
I think this is plausible, especially if you make the range of “somewhat longer” so big that it encompasses more than an order of magnitude of time, as in years—decades. It’s still not obvious to me, though.
Out of curiosity, where do you think people got the idea of going to the moon from? By your logic, since we never saw any animal go to the moon, how to do so shouldn’t have been obvious to us and it should have been extremely difficult to secure funding for such a project, no?
Yes, this is true. We note in a footnote that performing an anthropic update is similar to assuming an extra (virtual) success in the observation period, so you can indeed justify our advice of introducing such a success on anthropic grounds.
This is true, but if this level of similarity is going to count, I think there are natural “examples” of pretty much anything you could ever hope to build. It doesn’t seem helpful to me when thinking about the counterfactual I brought up.
I strongly disagree with this counterfactual and would happily put up large sums of money if only it were possible to bet on the outcome of some experiment on this basis.
Humans designed lots of systems that have no analog whatsoever in nature. We didn’t need to see objects similar to computers to design computers, for instance. We didn’t even need to see animals that do locomotion using wheels to design the wheel!
It’s just so implausible that people would not have had this idea at the start of the 20th century if people hadn’t seen animals flying. I’m surprised that people actually believe this to be the case.
This sounds implausible to me. I agree it would have taken longer to occur to people, but arguing that at some time people wouldn’t have figured out how to make helicopters or planes seems difficult to believe.
I’m curious why people would think this, though. Why is the possibility of flight and the basic mechanism of “pushing air downward” supposed to be so difficult, either conceptually or as a matter of engineering, that we couldn’t have achieved it without the concrete example of birds and insects?
I’ve seen this claim too, but I’ve also seen claims that historically obsession with bird flight was something that slowed down progress into investigations of how to achieve flight. On net I don’t think I make any updates on this evidence, unless I get a compelling account for why bird flight provides us with a particular insight that would have been considerably more difficult to get from another direction.
Edit: I also think the observations the Wrights are said to have made are rather superficial, and could not have been useful for much more than a flash of insight which shows them that a particular way of solving the problem is possible.
What would not be superficial is if they did careful investigations of the shape of bird wings, derived some general model of how much lift an airfoil would generate from that, and then used that model to produce a prototype to start optimizing from. Is there any evidence for this happening?
The difference is people have been trying hard to harness nuclear forces for energy, while people have not been trying hard to research humans for alignment in the same way. Even relative to the size of the alignment field being far smaller, there hasn’t been a real effort as far as I can see. Most people immediately respond with “AGI is different from humans for X,Y,Z reasons” (which are true) and then proceed to throw out the baby with the bathwater by not looking into human value formation at all.
I’m not sure why you would think this. The actual funding that goes into trying to do this is not that large; fusion research funding is maybe like $500 million/yr. The FTX Future Fund alone will probably spend on the order of this much money this year, for instance. Most of these proposals are aimed at one very specific way of trying to exploit the binding energy (turn hydrogen isotopes into helium and other heavier elements) and don’t consider alternatives.
I think this approach is basically correct because I don’t see any plausible alternative that anyone has come up with. Fusion is promising because we know it happens in nature, we can trigger it under extreme conditions, and there’s an obvious mechanical explanation for why it would work. The only challenge is an engineering one, of doing it in a controlled way.
If “humans are an untapped source of evidence for alignment” or any similar claim is going to have teeth, it needs to be coupled with a more concrete strategy about how we should go about extracting this evidence, and I’m not sure where I’m supposed to get this from the post. I would be highly surprised if anyone said “evidence from humans is irrelevant to alignment”; I think the actual reason people don’t go down this path is because they don’t think it’s promising, much like the people who don’t spend billions of dollars exploring possibilities of cold fusion.
Planes don’t fly like birds, but we sure as hell studied birds to make them.
I don’t think this is actually as clear as you might think it is. As far as I can see birds are useful for designing planes in only the most superficial way; that is “you need to be pushing air downwards so you can fly”. Birds do this in a way that is different from helicopters, and planes do it in a way that’s different from both. You don’t need to have seen birds to know this, though, because conservation of momentum gets you to the same conclusion pretty easily.
I don’t really see how any deeper study of birds would have helped you to better design planes. If anything, bird flight is so complicated that studying it could have made it harder to design planes because you’d try to replicate how birds do it instead of solving the problem from first principles, e.g. by trying to figure out which airfoil shapes would deflect an air current downward.
This is their current research direction, The shard theory of human values which they’re currently making posts on.
Sure, but why do they think it’s a promising direction of research? This is what’s not clear to me and this post hasn’t helped make it any clearer, though the shortcoming there at least partly lies with me due to my inability to understand what’s being said.
I find myself confused about what point this post is trying to make even after reading through it twice. Can you summarize your central point in 100 words or less?
If the title is meant to be a summary of the post, I think that would be analogous to someone saying “nuclear forces provide an untapped wealth of energy”. It’s true, but the reason the energy is untapped is because nobody has come up with a good way of tapping into it. A post which tried to address engineering problems around energy production by “we need to look closely at how to extract energy from the strong interaction and we need to check if we can squeeze anything out of that whenever new physics is discovered” would not be compelling.
If you come up with a strategy for how to do this then I’m much more interested, and that’s a big reason why I’m asking for a summary since I think you might have tried to express something like this in the post that I’m missing.
I don’t think I understand this post. Are you making the claim that “because some subset of human values are (and were selected for being) instrumentally convergent, we don’t have to worry about outer alignment if we project our values down to that subset”?
If so, that seems wrong to me because in most alignment failure scenarios the AI does actually have a terminal goal that to us would seem “arbitrary” or “wrong”. It only pursues the instrumentally convergent goals because they help the AI towards the terminal goal. That means you can’t bank on the AI not turning you into paperclips at some point, because it might judge that to be more expedient than keeping you around as another computer for doing research etc.
In addition, there’s the added danger that if the AI leaves you around, your inability to precommit to some strategies will always pose a threat to the AI’s totalizing vision of a universe full of paperclips. If so, it’s instrumentally convergent for the AI to eliminate or permanently disempower you, even if you yourself are currently aiming for the same goals the AI is aiming for, both instrumental and terminal.
For example, if I get smarter, would I stop loving my family because I applied too much optimization pressure to my own values? I think not.
This seems more likely than you might imagine to me. Not certain or not even an event of very high probability, but probable enough that you should take it into consideration.
That’s not quite true—there is at least the naive link that a higher equilibrium body mass leads you to expend more energy in daily activities even if you exercise the same amount as before. In my very naive model I assume these are directly proportional, but Natalia cites some better research that does a log-linear regression of calorie expenditure on equilibrium (I think? I didn’t check this part) body mass which seems to be more accurate empirically.
I think it’s unclear whether we had the link you mention in the past, too. We definitely had a correlational link: people who did hard labor and ended up exercising a lot every day took in much more calories, as we would expect, and they were generally not obese. However, I think my argument would work just as well in the past if you just applied the do operator on calorie intake per day and looked at the causal impact on equilibrium body mass, as I don’t think there’s evidence that there’s a big downstream link from calorie intake per day to exercise.
Yes, this summary is accurate.
I’m not sure who is “to blame” for the miscommunication but I suspect I simply was not clear enough in my top comment. Now it’s likely too late to clear up the issue for most readers as they won’t be following the developments in this thread.
I’d be interested to hear you elaborate on exactly how these attempts failed. What goes wrong if you look at vectors of the form vb+αvh for α∈R and vb,vh the vectors in latent space that you identify for brightness and height respectively, and then do some kind of manual binary search on α to eyeball the value that cancels out the brightness factor? Do you run into a collinearity problem when you try this, for instance?