"G" fits my own understanding best: "Not Okay" is a generalized alarm state, and the ambiguity is a feature, not a bug.
(Generally) we have an expectation that things are supposed to be "Okay" so when they're not, this conflict is uncomfortable and draws attention to the fact that "something is wrong!". What exactly it takes to provoke this alarm into going off depends on the person/context/mindset because it depends on (what they realize) they haven't already taken into account, and that's kinda the point. For example, if you're on a boat and notice that you're on a collision course with a rock you might panic a bit and think "We have to change course!!!", which is an example of "things not being okay". However, the driver might already see the rock and is Okay because the "trajectory" he's on includes turning away from the rock so there's no danger. And of course, other passengers may be in Okay Mode because they fail to see the rock or because they kinda see the rock but they are averse to being Not Okay and therefore try to ignore it as long as possible.
In that light, "Everything is Okay" is reassurance that the alarm can be dismissed. Maybe it's because the driver already sees the rock. Maybe it's because our "boat" is actually a hovercraft which will float right over the rock without issue. Maybe we actually will hit the rock, but there's nothing we can do to not hit the rock, and the damages will be acceptable. Getting people back into Okay Mode is in exercise in getting people to believe that one of these is true, and you don't necessarily have to specify which one if they trust you, and if the details are important that's what the rest of the conversation is for.
The best way to get the benefits of ‘okay’ in avoiding giant stress balls, while still retaining the motivation to act and address problems or opportunities is to "just" engage with the situation without holding back.
Okay, so we're headed for a rock, now what? If that's alarming then it's alarming. Are we actually going to hit it if we simply dismiss the alarm and go back to autopilot? If so, would that be more costly than the cost of the stress needed to avert it? What can we actually do to stop it? Can we just talk to the driver? Is that likely to work?
If that's likely to work and you're on track to doing that, then "can we sanely go back to autopilot?" can evaluate as "yes" again and we can go back to Okay Mode -- at least, until the driver doesn't listen and we no longer expect out autopilot to handle the situation satisfactorily. You get to go back to Okay Mode as soon as you've taken the new information into account and gotten back on a track you're willing to accept over the costs of stressing more.
"The Kensho thing", as I see it, is the recognition that these alarms aren't "fundamental truths" where the meaning resides. They're momentary alarms that call for the redirection of one's attention, and the ultimate place that everything resolves to after doing your homework and integrating all the information is back to a state which calls for no alarms. That's why it's not "nothing matters, everything is equally good" or "you'll feel good no matter what once you're enlightened" -- it's just "Things are okay, on a fundamental level alarms are not called for, behaviors are, and it's my job to figure out which. If I'm not okay with them that signals a problem with me in that I have not yet integrated all the information available and gotten back on my best-possible-track". So when your friend dies or you realize that humanity is going to be obliterated, it's not "Lol, that's fine", it's room to keep a drive to not only do something about it, a drive to stare reality in the face as much as you can manage, to regulate how much you stare at painful truths so that you keep your responses productive, and a desire to up one's ability to handle unpleasant conflict.
How should one react to those who are primarily optimizing for being in Okay Mode at the expense of other concerns
Fundamentally, it's a problem of aversion to unpleasant conflict. Sometimes they won't actually see the problem here so it can be complicated by their endorsement of avoidance, but even in those cases it's probably most productive to ignore their own narratives and instead directly address the thing that's causing them to want to avoid.
Shoving in their face more reasons to be Not Okay is likely to trigger more avoidance, so instead of trying to argue why "Here's how closing your eyes means you're more likely to fail to avoid the rock, and therefore kill everyone. Can you imagine how unfun drowning will be?" (which I would expect to lead to more rationalizations/avoidance), I'd focus on helping them be comfortable. More "Yeah, it's super unfun for things to be Not Okay, and I can't blame you for not wanting to do it more than necessary"/"Yes, it's super important to be able to be able to regulate one's own level of Okayness, since being an emotional wreck often makes things worse, and it's good that you don't fail in that way".
Of course, you don't want to just make them comfortable staying in Okay Mode because then there's no motivation to switch, so when there's a little more room to introduce unpleasant ideas without causing folding you can place a little more emphasis on "it's good that you fail in that way", and how completely avoiding stress isn't ideal or consequence free either.
It's a bit of a balancing act, and more easily said than done. You have to be able to pull off sincerity when you reassure them that you get where they're coming from and that it's actually better than doing the thing they fear their option is, and without "Not Okaying" at them by pushing them "It's Not Okay that you feel Okay!". It's a lot easier when you can be Okay that they're in Okay mode because they're Not Okay with being Not Okay, partially just because externalizing ones alarms as a flinch is rarely the most helpful way of doing things. But also because if you're Okay you can "go first" and give them a proof of concept and reference example for what it looks like to stare at the uncomfortable thing (or uncomfortable things in general) and stay in Okay Mode. It helps them know "Hey, this is actually possible", and feel like you might even be able to help them get closer to it.
or those who are using Okay as a weapon?
Again, I'd just completely disregard their narratives on this one. They're implying that if you're Not Okay, then it's a "you problem". So what? Make sure they're wrong and demonstrate it.
"God, it's just a little fib. Are you okay??"
"Not really. I think honesty about these kinds of things is actually extremely important, and I'm still trying to figure out where I went wrong expecting not to have that happen"
Or
"Yeah, no, I'm fine. I just want to make sure that these people know your history when deciding how much to trust you".
Curated.
This is a somewhat old post, but I thought it was underappreciated (other LW mods weren't quite sure why I wanted to curate it).
I think the core concept here was a crisp articulation of something I hadn't had a clear handle on, but I think is a key rationality concept. And as Zvi dug into the different examples, it was very illuminating how confused and jumbled my default intuitions about being "Okay" and "Not okay" were.
Distinguishing the psychological state of "things are okay/not-okay" from the reality of "are things 'okay' within some frame of judgment?" seems really important. This feels similar to Nate Soare's post about how to detach the attitude of 'conviction' from your epistemic assessment of "will Project X work out?"
It happened to be particularly relevant to me this week, as I was feeling a sense of 'things are not okay', and looking around, things in fact seemed 'not okay' in many objective senses. Nonetheless, being in "not okay mode" was making it harder to think sensibly.
I liked Zvi's exploration of how the concept of "Are things okay?" can get distorted at the group rationality level, where people might pressure people help them adopt a "I can be okay" stance by ignoring problems, or shuffling responsibility for them around, without noticing that that's what you're doing.
Longterm, I'd be interested in both:
A criticism I have of this piece, similar to some other Zvi pieces that do an exhaustive taxonomy, is that it's overly long and the individual examples are just given simple letters that get really hard to follow in the second half. (similar complaint on Simple Rules of Law, despite also liking that post a lot). I'm not actually sure how to solve the problem though.
oh dear. J is so familiar it was painful to read.
I expect to forward this post to people close to me, so we can better talk about Stress and Action without triggering all my Attack associations.
It seems quite important to separate OK mode from the state of the world (see also, Nates article on detaching the grimometer).
There is a problem, and that's ok.
I'm freaking out about it, and that's ok.
I have a response-ability o fix it, and that's ok.
Everything is ok, because what else could it be? . Reality is the way it is, and thats ok.
Ok, what do I do now?
It seems like not ok mode is the mode of 'get others to see my panic and create a plan, whereas ok mode is the mode of 'create my own plan'. Interestingly, this seems almost the opposite of your model.
One could imagine a 3 by 2.
It seems like the best place to be is yes in all 3, but it's probably better to be yes yes no than no no yes.
This is a beautifully articulated taxonomy. My first application is to addictions. At the top level, addiction "makes things okay" in a way that makes things very not okay.
Uppers can make someone feel like things are okay. This reduces stress until they wear off and the "not okay" can be seen more clearly, assuming the time spent "up" wasn't used to actually make things okay.
Downers can make someone feel like being "not okay" is okay. This reduces stress by taking away the onus of responsibility for a time.
Psychedelics can induce states of A or B.
Codependency is an addiction to trying to make things okay, in any sense of the word. Individually, it's a pathology. En masse, it's a driving force behind politics.
I think I've seen the 'J' dynamic before, but also it only "works" because others options are also in common use and so there's denyability of what's happening.
There is one other common case IMO: Monkey see monkey do. The "I have no idea what to say in the context and feel the need to say something/anything". Which has some in common with the empty platitudes one but may not be shallow as much as simply can't find words.
Cool stuff, like always a bit over my head. I’d be very interested to hear strategies about either getting “not okay” mode to “okay” mode, or for more accurately assessing when things are truly not okay (more more rarely than our brain states would suggest). It is my experience that in our modern world, and especially in the world of white collar work, this switch is flipped much much more than is necessary. When you get the email that you need to file some form that you probably should have known about, but in reality is no big deal, you definitely get sent into “not okay” mode both in the sense that there is a problem and you are technically at fault. On the micro level it is correct, but disproportionate in the reaction, and lacking context in that you’ll never be able to tie up the loose end hydra that afflicts us all. But that doesn’t stop the acute cortisol injection or the chronic stress! In my experience, the people who are most successful in these jobs ironically have the largest response to a stressor like that, and are obsessive about rectifying it. But one has to wonder the long term effects of that!
This is completely on the side, but I find your analogy "dreaded stress ball of uncertainty" strange. A stress ball is a small soft ball and squeezing it with you hands is supposed to be relaxing and make you less stressed.
Yeah, in hindsight the word order there is wrong, I'll think about fixing it, but also I never really "got" stress balls.
The answers to all these questions are complicated. While you may not be able to do everything perfectly every time, you're more likely to have the opportunity to do things better if you are operating from a place of greater focus. Your brain has been trained to operate under stress, so when you find yourself in an environment where you need to make difficult decisions, your brain tends to work harder than normal.
When people are dealing with negative self-talk, it makes them more likely to experience middle-nights sweats and hot flashes. When you are comfortable in your own skin, you recognize the middle-night sweats as the result of bothersome clothes. When you're in your own skin and experiencing hot flashes, freaking out isn't going to help.
All of this can be optimized for, or fought against. Acting rationally and with logic isn't "easy," but it's simple. We should all be able to acknowledge our animal natures. Once you realize that you're an animal, that's when things get interesting. You might be able to train yourself for optimal performance in stressful situations.
It's possible for your brain to start operating under Okay Mode issues... without falling into a cycle of stress. Be that as it may, you'll be much better in stressful situations, and it will be much easier for you to operate under Okay Mode if you take time out on a regular basis. This, at its core, is what it means to embrace "work hard, play hard," but on a somewhat larger scale.
Okay is a key concept for which we lack a good detailed handle.
There seems to be some sort of switch in brains, certainly in mine, that goes between modes of Okay and Not Okay. Being in Not Okay Mode is a giant ball of stress that makes it impossible to relax and demands attention.
When I hear talk that some problem exists, my first thought often has nothing to do with the substance of the problem. Instead, I feel instantly stressed and upset that someone has taken my perfectly pleasant Okay Mode existence, and declared that we are in Not Okay Mode.
Even if the real content of the problem is trivial.
There is an intense desire to get back to Okay Mode.
Even if that means one must be in Okay Mode by making Not Okay be Okay. Even if that means destroying all possible superior outcomes so that the bad outcome is now seen as acceptable.
Thus:
“Tell me everything is going to be okay.”
“Everything is going to be okay.”
What does that even mean?
It is a request for permission to enter or remain in Okay Mode. That requires unpacking.
When one says this, for what value of ‘everything’ and what value of ‘okay’ is this meant?
Does the future tense matter here? If everything is going to be okay, is everything also by implication okay now, or not? If we have to ask about whether things are going to be okay in the future, does that imply they are definitely not okay now?
Here are some possibilities that I think are central at least some of the time for what the question means.
A: Is ‘everything’ ‘okay’ in a way that is potentially meaningful in general but meaningless in a given context? The way that someone who experiences Kensho could say that everything was okay regardless of the current world state and value of everything, or that one might say that God has a plan? How central a feature of the basket of things called ‘enlightenment’ is the ability to label everything as ‘okay’ in a way your emotions will accept, and avoid the consequences of failing to do that?
B: Is ‘everything’ ‘okay’ in a way that is meaningful in a more specific context, but is a relatively weak claim meant to provide perspective? As in, yeah, I understand that sucks a lot, but look at the bigger picture and the things that actually matter are still ‘okay’. You spilled some milk, and the milk isn’t okay, but the relevant value of ‘everything’ does not need to include the milk, because the milk is not important. We still have plenty to eat, and/or anything up to and including it not changing the probability that AGI will be friendly.
C: Is ‘everything’ ‘okay’ in the sense that the dreaded stress ball of uncertainty will be resolved, and peace will be made with the ways in which things are definitely not okay?
D: Or, perhaps, is ‘everything’ ‘okay’ in the sense that you have put a Somebody Else’s Problem field around it so the whole thing can be ignored, which definitely resolves the stress ball brought about by uncertainty about whether something is in a state of okay or not? This is then assurance that, while things might be really awesomely terrible, no action is required or anticipated from you at this time – you don’t have to enter Not Okay mode, either in the sense of sending appropriate social signals for such a state, or in terms of actually fixing anything.
E: Or, similarly, a statement that no blame or punishment need be assigned for the current state of existence, and no scapegoat need be chosen.
F: Or, similarly, a request for an ‘I got this, you can treat this as having been handled’?
G: This could be a general question about whether to pay attention and/or change intentions towards future actions. Most people spend most time on autopilot, and this is a check to see if the autopilots need to come off. One could view D or also F as this applied only to oneself, whereas this is a request to confirm that all autopilots can continue to function normally. This is no one’s problem, and no one has to ‘get this.’ There is no need for anyone to suddenly be a person.
H: This could be a request to implicitly define ‘everything.’ There is something or someone or some group that is clearly not okay. Does it count? Do we care? Is it part of everything, or not?
I: This could be a request to implicitly define ‘okay’ in context.
J: This could be an attack. The respondent now has two choices.
If they affirm the state of okayness, several bad things happen to them. The two biggest are:
First, now it has been established that all the things that were previously wrong were ‘okay’ so they can’t go back and complain about them or blame anyone else for them later. By extension, other similar things that happen in the future are implicitly also okay, and perhaps the range of such things will now expand.
Second, they have affirmed that everything is okay, which means that if things turn out not to be, whether for them or for other people, that’s now their fault and they are blameworthy, and the claim will be made that we were in case F and then they didn’t deliver, but also other cases can also be invoked similarly.
Alternatively, they could deny okayness, and then different bad things happen to them.
First, they have now disturbed flow and made things socially awkward, to at least some extent. They have shown that they are a complainer, a person who is not okay with things, not comfortable. They are then seen as the aggressor, the one causing trouble and escalating. Bad improv. They will lose points.
Second, not only have they admitted they know there is a problem, they now have interacted with the problem. Oh no. All sorts of blame is now in play. All sorts of additional demands for explicitness can be made, and being explicit costs even more points. Exactly what is and isn’t okay, and exactly what counts as everything, are now among the questions that they might have to answer. So is how one might make things not okay.
Third, it is now on them to propose a way to make things okay again. Otherwise, who are they to claim that things are not okay without proposing a solution? And of course, if a solution is proposed, then the solution can be attacked and ridiculed.
Thus, the person asking if everything is okay is saying, either wave your right to complain and potentially be assigned blame if things go wrong, or make a stand and have it out now, while being marked as the one making things awkward and causing a problem, forcing us to be in Not Okay Mode, since otherwise everyone could have agreed to be in Okay Mode.
K: This could be a request for actual information about the state of the world. There is a set of things, everything, that can be in various states. How are those states? The better defined both ‘everything’ and ‘okay’ are in context, the more useful the question becomes. Note the distinction between asking about what social or other actions might be required, and asking purely for the world state.
L: This can be a demand for certainty. For choices and commitments – any choice and any commitment – to be made now in order to reduce the stress of the choices not having been made. This can happen for matters both great and small. If there are two possible futures, and both have big advantages and disadvantages, then one cannot be assured that any of the realms involved is “okay” because there is a trade-off and a different choice might get made. Thus, anything that could get worse or better or merely be different somehow is now Not Okay. Thus, even if there are large unknowns or there are otherwise huge advantages to not committing to a choice and no actual reason why a choice needs to be made yet, even if there is important information that will only be available later, optionality might have to be sacrificed on this alter.
M: It can be an accusation. You are pretending as if everything is ok, but I do not believe this, and you had better admit whatever it is that is not okay, or else.
N: It can be an assertion that someone should not objectively be Okay, and should instead be Not Okay, especially when combined with terms like “Are you sure that…”
O: It can be a social script for generic meaningless reassurance, which may be a request for an enactive statement rather than something descriptive. ‘Everything’ and ‘okay’ need not have any values at all, because to many, words don’t have meanings. Interpreting this as a claim that there exists a set of things that is being claimed to be in a particular state, or that will be in a particular state, would be a pure category error. Saying this merely builds momentum towards the Okay state.
P: This could be an establishment of dominance and submission. The exchange reminds both parties who decides whether things are Okay or Not Okay, granting power over the other’s emotional state, which can be used as power generally. And, of course, that power is now okay.
Q: This could be a debate about the ‘right to complain.’ Complaining can be about information and moving towards solutions or better understanding, but it can also (more frequently) be about scoring points, or about needing to feel heard, or punishing the scapegoat. Saying everything is okay in this mode is therefore an assertion that there is no right to complain, whereas saying that everything is not okay is both itself a complaint and a prelude to more complaining. Asking if everything is okay can be an invitation that this is a good time to complain (either something might be done, or you might get a sympathetic ear, or if you gotta do it all right let’s get it over with, or speak now or forever hold your peace) or could be the opposite.
R: Or perhaps is ‘everything’ ‘okay’ because actually yeah, actual everything is actually okay now?
There are a lot of these that combine multiple meanings, or can mean different ones in different ways. In some cases it would be reasonable to disagree with which meaning is the locally primary one, but this is an attempt to group them together.
Different contexts have different defaults and distributions of possible explanations. The D/F/G/K cluster of related interpretations is closest to my default generic assumption, with the differences between those three often being important. Also note that one can respond with a different type of response than what was requested, either intentionally or unintentionally, and make this clear through other words or even through tone, subtly (for example) transforming a ‘nothing need be done’ into ‘I got this’ or vice versa.
These groupings suggest a few categories.
A, B and R represent the cluster of universalized claims about all of existence. That seems central to these. In other ways they are mostly part of the second category.
C, I, L, N and O are treating okay as not stressing out about local concerns, and debating whether or not one has earned the right to be in Okay Mode or not.
D, F, G and K, are attempts to determine if action is necessary, and if so who must take what action, with varying levels of clarity creation versus emotional needs and blame avoidance.
E, M, P and Q are focused on assigning blame and on social roles.
J is an attack, which puts it in its own category.
That leaves H, which depending on what ‘okay’ means could be part of any of the middle three categories, sufficiently that I didn’t want to place it in any of them, but it falls into some combination of those.
The big two groups are those where the concern is ability to stress or avoid stress, and those where the issue is where it is necessary to take action or assign blame, with J as perhaps a special super-case in the second category.
A major motivating factor for creating this taxonomy is how easy it is for stress concerns to override non-stress concerns, and for the dominant dynamic to become what will allow people’s brains to give themselves permission to shift back into Okay. In this model, that is what drives any action taken, or any blame assigned.
What else can ‘okay’ and ‘everything’ mean, in what combinations? To what extent can we profit by tabooing the word okay, or most of these uses for it? What is the best way to reliably get the benefits of ‘okay’ in avoiding giant stress balls, while still retaining the motivation to act and address problems or opportunities? How should one react to those who are primarily optimizing for being in Okay Mode at the expense of other concerns, or those who are using Okay as a weapon?