Far-Future Commitments as a Policy Consensus Strategy

Let me start by saying I think I really should have said 50 years instead of 100. That seems to a big sticking point for people.

I haven't made myself clear. What I'm saying is create a contract to enforce the institution of a law. The government is the signatory, not any particular person. Governments can and do hold by agreements that past governments made, even if they disagree with their predecessors.

Now, sometimes they break those promises, and that's why I'm talking about creating a binding contract. As in, if you break the contract, a penalty occurs.

For example, let's say the government issues $1T of a particular financial asset. Those assets are basically inflation-adjusted perpetuities, except upon the institution of the delayed law, they cease to pay out. (Unless the law is instituted before the 50-year period, in which case they only cease to pay out after 50 years so that the perpetuity buyers aren't swindled.) If the law is not instituted, then they continue to be perpetuities.

Upon selling those "enforcement perpetuities", the government splits the ownership of land and buildings. So if you own a house, now you own the house itself, as well as a financial asset for the land underneath it. The land asset pays off a cashflow equal to the land value (if you're taxed $1000 on Jan 1st, the government pays you $1000 on Jan 1st). Now, the money raised from the enforcement perpetuities can be used to buy those land assets from all current landowners.

Maybe you'd call this an incentive mechanism, rather than a contract. Whatever, I'm not a legal expert. Whether that's the best solution or not, I don't know. Even if there's something wrong with the contract I just laid out, I would be shocked if there were not a way to do this. I'm sure there are plenty of creative solutions.

Obviously, that particular solution doesn't apply to all delayed laws. Different laws will require different solutions. For the voting system problem, make a constitutional amendment requiring a unanimous vote to overturn it.

I've just written a post on how exactly to enforce these contracts: Enforcing Far-Future Contracts for Governments [LW · GW].

Let me know what you think!

I see your point about constitutionality, but I think something such as the electoral system falls within its remit, which was my recommendation. But yeah, I would have no problem with extending the uses of the constitution, although I know that others would.

I agree that the financial product is more promising, because of its generality. The interesting thing is that even if they issue the asset with the intention of breaking the promise, they lose basically nothing in net present terms (because those cashflows 50 years into the future are discounted by 1+d to the 50th power). Thus it wouldn't be a mistake to do it (even while planning to renege). Since it's equivalent to issuing regular perpetuities in terms of how much money a given debt would raise, it's outright incorrect to call it a “bomb”. They don't have to pay $1T all at once at the 50-year mark, it's just that the present value of the debt will be worth that amount (assuming the same discount rate is the same as it was upon issuance).

Don't get me wrong, I get that “if you're explaining you're losing”. But LessWrong, of all places, should be an area to discuss such ideas, and not have it dismissed out of hand simply because “People are too stupid/lazy/disinterested to get it”. I don't think we should restrict our posts to 5 words or fewer. I don't think we should avoid novel ideas. And I don't think we should avoid politics.

The fact that everyone said Trump had no chance in the 2016 election shows that people basically don't know what they're talking about when it comes to political plausibility. No offence meant. I just don't see any value in such guesses, when they're so often wrong, and they stifle ideas.

Why is it that mathematicians are confident about their results? It's evident that they are highly confident. And it's evident that they're justified in being confident. Their results literally hold for the rest of time. They're not going to flip in 100 years. So why is this the case?

Basically, there are a few stages of belief. Sometimes a belief is stated on its own without anything else. Sometimes a justification is given for that belief. And sometimes it's explained why that justification is a reliable indicator of truth. Now, you may think you have infinite regress here, but that's wrong in practice: You eventually reach a point where your justification is so trivially obvious that it almost feels silly to even list those justifications as assumptions in your argument.

Voting systems have been figured out to the level of standard mathematical detail (i.e. belief + justification). And my methodology post that I linked to you is explaining why justifications of the form they use are unambiguously correct in the world of policy. (Again that series is not finished yet, but the only roadblock is making it entertaining to read, I've already figured out the mathematical details.)

So to me, arguing against a voting system change is like saying “Maybe there are a finite number of primes” or “Maybe this table I'm resting my arms on right now doesn't actually exist”. I.e. these really are things that we can, for all intents and purposes, be certain of. And if you're not certain of these basic things, we can't really ever discuss anything productively.

It's not a matter of the Dunning–Kruger effect; it's that experts understand these problems well enough. You can find professors who specialise in voting theory and ask them. Ask them “Is there any chance that replacing the current presidential voting system with any of the most promising current alternatives will be a mistake in 100,000 years?” The amount of time is totally irrelevant when you understand a problem well enough. One plus one will always equal two.

Conversely, AI safety's whole problem is that we don't have anything like that. We have no confidence that we can control these systems. We have proposals, we have justifications for those proposals, but we have no reason to believe that those justifications reliably lead to truth.

To be clear, I'm not saying every policy problem is solved. But some policy problems are solved. (Or in the case of voting theory, sufficiently solved as to far outperform the current system, and we know that there is no unknown system will blow our current proposals out of the water due to Arrow's impossibility theorem.) And establishing some of those policies is difficult because of short-term incentives. This delay tactic is a way to implement that specific subset of policies, and only that subset.

Denying this would require you to think that no such policies exists. Which would commit you to say, “Hey, maybe the Saudi Arabian policy of cutting off a child-thief's hand shouldn't be revoked in 50 years. Who can say whether that'll be a good policy at that point?”

There are cases where there's a consensus among lawmakers that it would be good to do something and there's current opposition that can be circumvented by writing the law into the future

This is exactly my point. It's actually the entire point of the article. Everything else was simply trying to describe this problem and give examples so that it's easier to understand. Whether anyone agrees with the examples does not matter. The article literally emphasises this point: “The examples I’ve given are simply for clarity.”

Actually addressing your specific example is not strawmanning.

Should I have been more general than saying “Republicans and Democrats” interests are the bottleneck? Sure, it's the people in control of the actual vote that matters. But those details are entirely incidental. It's kind of like pointing out a spelling error, and making a huge deal of it.

Now, I do want to hear this. What about your overgeneralisation that “Anything that calls itself an X acts like an X”? Do you not concede that you were straw manning me in that instance?