Just this guy, you know?
Dagon
Karma: 11,593
if I were perfectly rational agent …
Yeah, “perfectly rational” implies consistency over time, which is the whole question you’re struggling with.
if you keep adding strict prohibitions onto your future self, you are limiting all your future options
Right, that’s kind of the point, isn’t it? You don’t trust your future self to be consistent with your present beliefs, so you constrain it. Note that there are current-self tactics based on the same principle: you might not trust parts of your decision apparatus to make food choices, for instance, that other parts of you prefer, and therefore don’t keep junk food at hand.
Most older people I know (including myself), recognize that their younger selves were jerks, or at least confused about a lot of things. They often regret commitments made previously (and often agree with them, but in those cases the commitment isn’t binding, as they like it anyway).
I’d recommend not framing this as a negotiation or trade (acausal trade is close, but is pretty suspect in itself). Your past self(ves) DO NOT EXIST anymore, and can’t judge you. Your current self will be dead when your future self is making choices. Instead, frame it as love, respect, and understanding. You want your future self to be happy and satisfied, and your current choices impact that. You want your current choices to honor those parts of your past self(ves) you remember fondly. This can be extended to the expectation that your future self will want to act in accordance with a mosty-consistent self-image that aligns in big ways with it’s past (your current) self.
This framing is consistent with most of your concrete suggestions—those are reinforcing the importance (to you currently) of these values, and the memory of this exploration, documentation, and thinking about why you currently care about future actions will inform your future self’s values. It’s not a contract, it’s persuasion.
Fully on-board with the annoyance at this equilibrium. I don’t see a better way, unfortunately, with the information and motivation asymmetry between software employers and employees, both of which have large variances in quality.
I’ve focused on the technical and social/team aspects of software development as a (very) senior IC, rather than as a manager in title. Even so, I’ve been deeply involved in hiring, organizing, motivating, and aligning teams for a number of large projects. I’ve found a very strong correlation between the signaling of “work ethic” in terms of energy and hours and the actual performance and impact of an employee. Like all heuristics, it’s nowhere near 100%, and it’s sad that there’s no easy way to identify the exceptions. Sad as it is, it’s true—as an employer of software engineers, I would prefer not to hire part-time.
Which means the expected-productivity curve for employers is nonlinear, so there’s no reasonable way to make the pay/effort ratio constant.
100% (or more—this justifies hyperbole) on applying “work ethic” to other aspects of life. This difficult tradeoff of motivated effort on behalf of others applies to housekeeping, care for partner/children, some parts of hobbies, and a lot of other things. It’s not work-for-money, it’s work-for-others-preferences.
I love this idea! I don’t actually like videos, preferring searchable, exerptable text, but I may not be typical and there’s room for all. At first glance, I agree with your guess that the overview/intro is more value per effort (for you and for consumers, IMO) than a deep-dive into the code. There IS probably a section of code or core modeling idea for each where it would be worth going half-deep into (algorithm and usage, not necessarily line-by-line).
Note that this list is itself incredibly valuable, and you might start with an intro video (and associated text) that spends 1 minute on each and why you’re planning to do it, and what you currently think will be the most important intro concept(s) for each.
he point is that wanting a part-time job without having a really good excuse is a bad signal of one’s work ethic. What else could be a good excuse?
One part of the declining societal work ethic is it’s quite a bit more common to decide to work less, without needing much of an excuse. I know a LOT of people under normal retirement age, who describe themselves as “semi-retired” or “working enough to keep my head in the game”, rather than optimizing for career and future earnings potential. Note that this is mostly long-term software devs who’ve amassed quite a lot of savings, and I ALSO know many who are near or past normal retirement age and need to keep working for the money. A lot of motivation isn’t about work ethic, it’s about transactional optimization.
The tradeoff of “pretty good compensation” for “more time/energy than I’d like to spend” is pretty rampant. That part isn’t about sexism or having a good excuse, it’s just the bundling that seems to work for most employers.
Big feels on the “This sucks, it’s gotten worse, but objectively I’m so much better off than the VAST majority of past and present humans that it feels petty and unjustified to complain” front. Your framing of “in theory it shouldn’t be so hard, but in practice, it is” is an excellent summary.
I suspect it comes down to the fact that all perception (of the physical and of the social world) is comparative—you notice things and evaluate them only and exactly in comparison to expectations. And since those expectations are mutable and imaginary, we are noticing how shitty things are compared to what our rosy memories and what media shows us. Comparing to a 12th century peasant farmer would make us happier, but that’s far less available.
(the sexism comment doesn’t match my experience in software dev in England and in the US—there are almost no good part-time jobs for women, only adjunct or career-limited positions, which could go to men if they wanted, but they generally don’t. Very flexible schedules for childcare or anything else are pretty available to both men and women, regardless of who’s primary caretaker. This likely varies widely, but I’d be shocked if you had to provide proof of divorce to get accommodation).
Cancer (and other) vaccines seem pretty cool.
[ speculative; I see some indications of this in my personal experiences and those I talk with, but I have no idea how prevalent it is. ]
I’ve long struggled with motivation and akrasia—since before the Internet was a thing. It’s gotten slightly better and slightly worse a few times, more based on my age and situation than (I think) on external factors. I AM more open about it in the last decade or so than previously, mostly because others have started talking about it, and I’m (slightly) less afraid that I’ll be held accountable for my failings more forcefully when I admit it than when I hide it. It may be that it’s harder for me to overcome it when I’m not trying to show the world it doesn’t affect me.
I think this social acceptability and visibility plays a MUCH larger part in people’s behavior than is often admitted. The tension between “accept and love people uncondionally” and “encourage cooperative/conformist behavior, even if that’s not someone’s natural default” may have no great equilibrium point. It’s definitely shifted AWAY from “conforming enables cooperation”, which seemed to be the common viewpoint in my youth and young adulthood, TOWARD “be yourself, even if it inconveniences others. Anyone who sees a downside is a bigot (even if they are only pointing out direct risks and suggesting mitigations that don’t seem to them to deny someone’s identity).”
It’s massively exacerbated by the changes in media and scalable individual visibility—what we see as “normal” is the tail of the distribution that is shown to us over and over, and there are way fewer role models for conformist success in the mix.
This applies to a whole lot of topics, but “work ethic” is probably one that seems most impactful. I really do look forward to seeing if anything remains of a world where a large majority is not willing to be drones supporting the general equilibrium (including vast wealth for the oligarchs).
In my own mind, there’s a very large tension between “nobody should have to do that” and “boy, it’s going to suck if nobody actually does that”. This applies to picking berries, office work, doing laundry, raising children (the MASSIVELY time-consuming and unpleasant parts, especially), and almost everything else. I don’t have a good answer, and I don’t have much expectation that “AI will save us” will happen at all, let alone happening before societal meltdown.
I think this would benefit from a crisp definition of “consciousness” and of “simulation”. THEN you can clarify your question to one of:
MUST consciousness be simulated, because it can’t exist in a base-level reality.
DOES my consciousness happen to be simulated, even though it’s feasible to exist in a base-level reality.
CAN consciousness exist in a simulation, or is it only conscious in the base-level reality, with the simulation being some sort of interference layer.
I haven’t seen good enough definitions of either thing for these questions to make sense. Most conceptions of ‘simulation’ are complete enough that it’s impossible to determine from inside whether or not it’s a simulation, so that would lead to “with that conception of simulation, with the consciousness that I’m experiencing, it is untestable and unimportant whether it’s in a simulation”.
It’s not actually iterated, it just allows communication before the final agreement (or not). Simply assign values for each player for s and t, and you can calculate various equilibrium outcomes. For fun, it’s not guaranteed that either party knows the other’s utility function over (s,t), so you can make each player assign a distribution of utilities for the other, and figure out the optimum for each. Then negotiation can reveal some of it (or can fail and lead to sub-optimal outcomes).
I can only speak for myself, but I downvoted for leaning very heavily on a current political conflict, because it’s notoriously difficult to reason about generalities due to the mindkilling effect of taking sides. The fact that I seem to be on a different side than you (though there ain’t no side that’s fully in the right—the whole idea of ethnic and religious hatred is really intractable) is only secondary.
I regret engaging on that level. I should have stuck with my main reaction that “individual human conflict is no more likely to lead to AI doom than nuclear doom”. It didn’t change the overall probability IMO.
I’ve played a lot of poker, both socially and in cardrooms. I would probably not join a game that needs legible commitment mechanisms. If it’s not sufficiently high-value to succeed on it’s own, I probably have better things to do most of the time.
Mostly my recommendation is to find/create games that people like enough and feel safe enough to tell each other their expected attendance. Also, get good at short-handed play—it’s useful if a few people drop out, but even more useful to be able to play while some people are still en-route.
Not on my profile page, as far as I’ve been able to tell. Not important enough to dig around GW. I think I’ll start just saving local copies of some of my comments, and decide whether to turn them into a shortform if I notice they’re deleted.
Would it be simple for the system to send me a PM with the contents when a comment is deleted by a post author (either directly or by deleting the post that contains it)?
Ehn. Kind of irrelevant to p(doom). War and violent conflict is disturbing, but not all that much more so with tool-level AI.
Especially in conflicts where the “victims” aren’t particularly peaceful themselves, it’s hard to see AI as anything but targeting assistance, which may reduce indiscriminate/large-scale killing.
I haven’t noticed it (literally at all—I don’t think I’ve seen it, though I’m perhaps wrong). Based on this comment, I just looked at https://www.lesswrong.com/users/review-bot?from=search_autocomplete and it seems a good idea (and it points me to posts I may have missed—I tend to not look at the homepage, just focusing on recent posts and new comments on posts on https://www.lesswrong.com/allPosts).
I think putting a comment there is a good mechanism to track, and probably easier and less intrusive than a built-in site feature. I have no clue if you’re actually getting enough participation in the markets to be useful—it doesn’t look like it at first glance, but perhaps I’m wrong.
It does seem a little weird (and cool, but mostly in the “experiment that may fail, or may work so well we use it elsewhere” way) to have yet another voting mechanism for posts. I kind of like the explicitness of “make a prediction about the future value of this post” compared to “loosely-defined up or down”.
I liked it, but probably don’t want it there all the time. I wonder if it’s feasible (WRT your priority list) to repeat some of the site feature options from account settings on a “quick feature menu”, to make it easy to turn on and off.
If I make a comment, then the author deletes the post, is my comment lost as well? I’m pretty sure I don’t actually need a record of things I’ve said, and that my comments aren’t valuable enough to justify extra effort, but it kind of bugs me.
Has anyone written the scraper or script to save a copy of everything you write on LW, so it’s available even if it gets deleted later? If it archived the post and comment thread you’re responding to, it would be all right.
My suspicion is that the legibility and exactness of financial system won’t be repeated for other kinds of “social credit”. The keys that make the financial system work are:
Fungibility—a quantity of credit from A or B is equally valuable. This applies to money or “simple goods”, but does not apply to trust, liking, respecting, or fearing.
Fixed quantity—financial values are neither created nor destroyed (by normal transactions—there are underlying system actions that do change totals). Trust, liking, fear, and respect are not conserved in transactions.
Transferability—the previous two properties make it pretty straightforward to exchange at scale, leading to large benefits from specialization and trade. It’s not clear how status, power, or friendship can be made to scale in those ways.
Really interesting exploration, I’m glad for it. I would warn about applying it to non-randomly-selected pairs—most claims of (non-)causality are not taken equally from a distribution of functions.
Butlerian Jihad? Sign me up!
I’m not sure I follow why Aumann’s agreement theorem is relevant here—the survey does not include any rational agents, agents with mutual knowledge of their rationality, nor agents with the same priors. It makes sense to one-box ONLY if you calculate EV by that assigns a significant probability to causality violation (your decision somehow affecting the previously-committed Omega behavior).