Epistemic Status: Love the player, love the game
Also consider: Playing on Easy Mode
Raymond Arnold asked me, why do you insist on playing in hard mode?
Hard mode is harder. The reason to Play in Hard Mode is because it is the only known way to become stronger, and to defend against Goodhart’s Law.
Strategies that work in Easy Mode won’t work in Hard Mode.
The key idea of Hard Mode is to keep your eyes on the prize. You know exactly what you want. You can’t munchkin your way to getting it. Once you start aiming to make a number go up, or get a check in the right box, you have lost sight of the thing you actually want. Proxy measures lead to failure; your value is fragile. That number correlates to what you want, but only insofar as you’re aiming for the goal and not the number. If you break the spirit of the exercise, all is lost. Your values have been hijacked. If you fail to develop skills along the way, you have missed the point, because the game has no end.
I
Consider playing guitar in Rock Band. You must choose whether to play in Hard Mode. If you do, you will fail a lot. You will play the same songs over and over again. Tricks that rely on there only being so many notes, or going at a relaxed pace, collapse. Eventually, you learn new techniques. You get better. You play on expert, your fingers get sore and you smile as you sing along.
II
You have a test in a week. You ignore it. You’ve asked questions based on your curiosity, to resolve your confusion. You study what is interesting to you, and what you feel would help you in the future.You focus on learning key principles, knowing you can derive what details you need later on. When the test comes, you work to figure out the answers. When you get the test back, you know how much you have learned. A year later, you remember everything, and build upon it.
III
You prepare for a tournament. You seek out the toughest opponents to help you prepare. You stop to criticize each other’s technique and point out every little mistake, no matter how irrelevant to the ultimate outcome of the practice match. You ask why and how you made that mistake. You do the same when you learn something new in a surprising way. You focus on the fundamentals, and don’t worry too much about exactly who you are up against this week. During the matches, you remember every tough decision and every mistake, so you can train again next week.
IV
You start a website writing articles devoted to the things you care about. To monetize it, you sell advertising through Google. It does not pay much at first. You keep at it, attracting a small but devoted readership. Some were already your friends, others soon join them. You look at what resonates so you can get feedback, but are careful not to take actions designed to maximize page views. Over time your writing improves and you learn much together. A community of sorts arises. You don’t quit your day job, but you teach others what you have learned.
V
(Spoilers for the excellent Groundhog Day)
You are stuck in a small snowed-in town, caught in a time loop of unknown origin. At first you have fun doing absurd things, but then you buckle down. With unlimited time, you decide to develop the skills and knowledge to give everyone a perfect day. You learn to play the piano, you read great literature. You listen to and remember the stories of everyone in town, and grow fond of them, learning what your opportunities are to engage in small acts of kindness. At the end of the day, after sufficient iterations*, you know you will be proud of your accomplishments, because you’ve made yourself and the world better, and you just might impress the hell out of your crush. If the loop continues*, you can do it again.
* – Results not guaranteed. You are unlikely to be in a movie. Local maxima may or may not be sufficient.
Interlude!
VI
Your help your friends move. With time and practice, your group of friends gets quite good and reliable. Any time, day or night, if someone needs to relocate, you’ll all be there, no questions asked. You call yourself the Midnight Movers. Most of your stuff arrives safely at its destination with a minimum of fuss, and you order everyone pizza. Your group draws closer together, and eventually tries going into business together on an unrelated matter.
VII
(Medium spoilers for The Good Place, highly recommended, skip this if you haven’t seen it yet)
You have an idea for a television show about a group of strangers who arrive in a mysterious place that plays by very different rules than our reality. You figure out exactly how this place works, plot everything meticulously, and lay out mysteries for the characters and viewers to uncover slowly over time. You use flashbacks that parallel events to examine and deepen the characters. Your production values are top notch and you produce great television. Your show is not a smash hit, and you know exactly where you are going with all this, so you don’t waste a minute, keeping your seasons short. In the end it all fits together, and the journey was still pretty great on second viewing.
VIII
Despite the terrible odds against you, you decide to strike out and open a fine Italian restaurant. Your dishes are sublime, but you soon learn that is but a small part of a successful enterprise. You must hire quality staff, arrange logistics across many suppliers, draw in customers and much more. Each step of the way, while ruthlessly keeping costs in check, you answer the question of what you would want your place to be like, and evolve your menu to offer a small selection of the best things you can affordably make. You get to know your customers by name. A casual observer would think you almost dislike money, as you can barely tell from the outside that there is a dining establishment there at all. Slowly word spreads among the cognoscenti, who show up and order the wine. Business is good enough. You get to keep working on perfecting your art.
IX
You have something to prove. To yourself.
Refuse to hire a cleaning service, ever, even though it’s totally worth it.
Tell the job interviewer your true strengths and weaknesses. If they don’t hire you, you didn’t want the job. Keep looking.
Find people to come to your meetup by promising them interesting intellectual discussions and a community devoted to truth.
Refuse to pirate music, television, movies, software, even when the owners are being kind of a dick and won’t sell it to you.
When you are in power, respect the minority even when you have the votes, don’t change the rules to pass the laws you want. Strengthen free speech rules and don’t silence those you disagree with, lest they do the same to you.
Write carefully as you fill out the forms. They might look like bureaucratic nonsense, and no one is likely to ever read them, but you should get this straight and cultivate good habits.
Learn to speak the language.
At your meetup, welcome challenges to in-group principles, so your group will be viewed better and feel more welcoming to and attract more members who seek the truth more than they demonstrate membership in the in-group.
Show other people what to do by example.
For demo day, you show what your system can do, and hope that you can keep building it. For that you need funding.
X
Dismayed by terrible things, you devote your life to the promise of artificial general intelligence. You discover that contrary to your initial beliefs, not only is creating AGI not easy, most versions of it kill everyone and destroy all utility in the universe. Explaining this is super hard. None of your explanations work. No one understands the danger. You set out to teach the world rationality, hoping this will cause them to see the potential dangers, with limited success. You write a book that’s silly but gets you exposure. You keep writing. Machine learning accomplishes more things and starts to get more funding. People start to come around to AI being dangerous, but mostly for the wrong reasons, so you don’t expect anyone to take the right precautions, and fear the world is doomed. You think that when they arrive, it will be far too late to correct for safety problems later.
XI
You are at a meeting to arrange educational services for your son. You know that the only thing that matters is what is written on the education plan. Whatever is in that document is what will count. Still, you cannot let the numerous falsehoods and stupid things pass, even though you realize that if you just play nice, they are going to put down on the piece of paper the thing that you want on the piece of paper. If you keep arguing, you risk getting nothing. Luckily, you think better of it in time. You go into easy mode. They write the words you need on the piece of paper. You sign it. You walk away happy.
In conclusion:
Pingback: Play in Easy Mode | Don't Worry About the Vase
Pingback: Paths Forward on Berkeley Culture Discussion | Don't Worry About the Vase
I think I can rephrase my question as “what are the specific goals you have your for your blog such that sharing relevant posts to LW, or various subreddits, works against those values.”
(Among other things, I think it is *bad* for people to be operating in isolated blogs that are harder to remember and build a common body of knowledge out of, at least for areas where there are existing places for that body of knowledge to build.)
((EASY mode is posting on FB, and I can respect your reasons for not wanting to do that))
I think this is a legit question for LW but that reddit is not a build-body-of-knowledge place.I am actively considering posting at least some things to LW but not to reddit, although of course others are free to post wherever they like.
There are issues of distorted incentives and information loss, in particular my wanting to feel free to do things I can’t do at LW and not feeling like that is expensive, but I don’t think those are big enough to be true objections.
My first true-objection concern with posting on LW, right now, is that it would shift the conversation away from here and into those locations, or splinter the conversation, which would be bad – I want the discussion to take place here, in this form, and have been very happy with how that is going.
My second true-objection concern is simply that LW is not currently well-functioning. If there is a LW 2.0 alpha or beta, I have not been invited, and 1.0 hasn’t had an article promoted in months, so I don’t think posting there would actually accomplish the no-splintering goal. In fact it would do the opposite, since they would now be in a sea of other things. Also, you can still find my stuff from LW due to Anthony’s link posts, which are 100% to include all my posts since the revival.
Right now, my instinct is that trying to cross-post would actually make things *harder* to follow in the long run rather than easier, given the current base state.
If whoever currently promotes articles on Less Wrong ever wants to *promote one of my posts to main* I will happily post that article to LW for them so they can do that.
If LW 2.0 is released, or a similar other website reaches critical mass with my participation, I will be happy to port the things that are sufficiently good and evergreen to be worth porting. To make it clear that I would actually do this, I would consider it sufficient for Sarah or Ben to explicitly email me and say, “Zvi, things are good enough now that we should port everything, I am porting everything and you should also port everything.” while that person in fact ported everything.)
Note that I have considered Arbital, because I actually want some of its functionality, and would be willing to consider engaging with it if it were a going concern, but doesn’t seem to get us anything for me to be the only one there, or the only one visibly there (it’s entirely possible that pages are still being written behind the scenes without telling anyone, some of which were quite good).
Basically I do think your concern is legitimate, but I don’t see a good way to address it right now. One of my to-do things for the blog is to create a link collection of the evergreen stuff in the rationalist sphere, because it’s one of those someone has to and no one else will situations and I think I would enjoy the walk – plan would be first to do best-of collections for the blogroll, then have a unifed best-of, plus walking through the LW archives and sequences and such. The Encyclopedia Rationalia not existing is indeed sad but seems not that expensive to create.
Fair.
*Right* now in particular the winning move seems to be “wait a few weeks or months for LW 2.0 to enter open-beta.” (At least, this is more winning than “post things to LW.”)
Agreed. I did see your post to my wall just now, so I have a better idea of what is planned. I hope they make LW 2.0 work and will help with that project to the extent I think it has a chance of working (it will also help if they coordinate with me, so far I’m mostly in the dark).
You’re probably already aware, but deluks updated the List of Blogs on the LW Wiki, so now there’s quite a lot of sites linked:
https://wiki.lesswrong.com/wiki/List_of_Blogs
I’m sort of on the fringes of the community, so maybe I don’t count as much, but I greatly prefer reading individual blogs over needing to dig through LW to read what I want to read. I like sequences. I also find it helpful when I become familiar with a particular writer’s thoughts, perspectives, and even biases. When someone I’m familiar with changes their mind on something.
I haven’t gone to LW regularly in years, and it feels really hard to get caught up on where the conversation’s gone from the site alone. Maybe my perspective would be different if I were constantly immersed in the conversation—i.e. had a constant headache from spending hours every day reading all the new and popular stuff. I’ve done that before and I think I did a good job of it. I felt informed and engaged and damn it I still missed the pre-2007 econ/thinking person blogosphere. Once I decided to phase out being on FB much, I felt better overall, and confused or obsessed in more genuinely constructive ways.
I do now notice that I can filter a particular user’s top level LW posts, which sort of simulates that user having their own blog. That gives me the idea of bookmarking a few key users’ LW submitted feeds on my phone. But on LW we still don’t have the benefits of tags generated by that user, or a timeline of post just for that user, or any kind of about or meta pages, which most basic blog setups would support.
At the moment, if I wanted to follow the most important parts of the conversation in the rationalist community by only looking at only one website, I’d just check in with Slate Star Codex and wait for Scott or commenters to link to relevant stuff.
I would change my mind about LW if it became less like a giant content feed and more like a community of individual blogs (along with the extra karma features if you really want). I guess Medium is sort of like that, but it’s still too much like a social network. In my heart I probably want another LiveJournal with a better UI. I know most of the rationalist commnuity LJ people moved to Tumblr. I haven’t adapted.
From a writers’ perspective, it’s more extreme. I’d strongly prefer to write in my own space. The opportunity to post on Less Wrong yields the pressure of writing an op-ed piece with none of the specialness. It’s hard enough to think things through and communicate those thoughts in writing. (Harder for some of us.) It’s harder to do that worrying, even in the background, how many freaking points it’s going to garner from people I’ve never met, which will be displayed prominently in a green circle along with the title of my post.
Nah, I’ll think, Let’s pass on wasting three hours writing someone people might not even find helpful.
And on a bad day: Maybe what I have to say, even if I’m willing to spend two hours trying to improve it, isn’t important. Maybe I’m insignificant and useless and don’t have anything to add to the community. And you know, of course that’s my fault and my problem, that I’m making it about me! So maybe I should just go ahead excuse myself again completely, as I’ve always done.
FWIW, Dreamwidth seems to be the most common LiveJournal replacement, though I don’t know that it has any better UI.
@sniffnoy: I’ve heard that about Dreamwidth but didn’t see the incentive to consider moving there.
@olimay: Tumblr, as I note elsewhere, is something I may have underinvestigated as a place where good people are and perhaps good content, but the design is pretty awful for discourse – trying to have this conversation there would be not great, whereas at LJ it would be as good or better as here, and I only abandoned LJ because others had abandoned it (plus the ownership situation, so I don’t fault anyone for folding there).
For now, I look forward to seeing what LW 2.0 can do.
Pingback: Rational Feed – deluks917
I think my problem with this outlook is that it kind of strikes me as fantastic. At least to my reading, these posts are about swimming against versus swimming along with systemic incentives. And the fantasy is that you as an individual can resist systemic incentives, that if enough people do that they’ll go away. That’s not true though, the entire point of systemic incentives is that they’re incentives and systemic. At best you can create an insulated bubble that’s locally immune to incentives, in the same way that physics permits you to temporarily defy gravity through sufficient use of energy.
But unlike gravity, systemic incentives are implicitly created by people, and you can do much better than resisting them. If you always resist systemic incentives, you’ll lose. If you sell out entirely you won’t be yourself anymore. If you just teeter on the edge of resisting as much as you can get away with…you live a precarious existence carefully calibrated to allow survival in a harsh environment. The only way to defeat systemic incentives is to change the system, and you’re not going to do that by providing a particularly authentic Italian restaurant with mediocre profits.
I put #11 there to make it clear that I was not saying to play in hard mode all the time. I agree that saying you should always be in hard mode is fantastical, and I tried to not stack the deck.
The point of providing a particularly authentic Italian restaurant is not that if you do it, it will lead to Olive Garden going out of business or change the world. It won’t (nor would I want it to, people forget how bad the things it typically replaces are, and my friend Paulo would be deeply sad). But if you respond along local gradients, you will end up in a place that isn’t authentic, so if you care about that – if that’s why you started the place – you need to fight the temptation and act in ways that look locally like burning money. Empirically, (almost) everyone who succeeds at this particular goal, does exactly that.
Do I have the fantastical goal that we can overturn the system this way? I think more people showing the way by example would help, but I agree that it’s unlikely to get it done.
One thing I thought about talking about is the concept of slack. Playing in hard mode requires slack – you need to be not living on the edge to a sufficient extent that you can afford to provide for the not-short term at all, and doubly so for non-obvious ways; this isn’t only a story about money or even time, but many things.
There are two ways to change the system: You can aim to change the system for a local culture like ours, or you can aim to Change The System, full stop. For the first one, I think it can be done. Again, you need to pick your battles carefully, and the first step is giving people a common vocabulary to talk about the problem, then the second is to provide counter-incentives to lower the slack costs. For the second one, the job is much harder. Not only are people in general orders of magnitude harder to reach, you have a very limited payload complexity you can hope to deliver. I don’t have plans to go that route. I do know of some simple laws/principles that would at least help, if I got hold of the rule-making hat, but I think that’s the wrong target to try and hit until at least we have local proof by example.
Easy mode: you work at a company doing easy things in a niche they control. You are among the smartest 10% of people and have worked there for several years, so you can produce enough to stay employed even by working just 15 hours a week. You’re bored at work but have a lot of time for your hobbies, and you get rich slowly but steadily.
Hard mode: you take an internship at a company doing something that people say is impossible in a highly competitive area. You’ve tried to do something like this before in a crappier company and failed. There’s a chance you are in the least smart 10% of the company, so you’ll have to work crazy hours just to survive. You don’t know how it will work out.
Apropos mostly of the Uplift video: last week I recorded a new song (well, pair of overlapping songs) that you might like: https://spintunes.bandcamp.com/track/be-kind-call-names-ft-melissa-phillips-shadow
(The right venue for this song would be as part of a musical, but the musical it’s for doesn’t yet exist.)
Now write easy mode in not strawman text.
Somewhat absent here is the idea of stakes. I think it is telling that in the one example where actually achieving a good outcome is personally important, the post suggests switching back to “easy mode”.
Numbers IV and VII, about monetizing a website / restaurant, are written from the perspective of someone who doesn’t actually need money and is financially set. If you had a family member with medical bills, or even student debt to pay down, I bet the post would suggest going with easy mode as well.
Numbers II and III are written form the perspective of someone who gets to keep playing no matter what. If you were one low grade from failing out of school or one loss away from relegation, than again I suppose the post would suggest going to “easy mode”.
That’s why I kind of think that easy vs. hard might be the wrong frame. A better one might be invest vs. succeed. When the thing itself doesn’t really matter, you pursue strategies with a lower chance of success but that build up skills that are useful for later. When the outcomes actually matter, you deploy the plan most likely to succeed.
Also, and somewhat of an aside, I think IX seems to have a number of pet peeves rather than actual easy vs. hard mode illustrations.
(IX was the hardest to write, and didn’t quite fit the format, but I felt I needed to put it in anyway. Agreed it was a little awkward.)
Your points on the examples are right, and in fact I intentionally tried to keep the playing field ‘fair’ in the sense that Hard Mode was not a dominant choice anywhere (other than IX, anyway). I intentionally tried to give easy mode better material/conventional payoffs to avoid stacking the deck.
Invest vs. Succeed is closely related to Easy vs. Hard, and they correlate highly – often you’re doing both. I think they’re both important. What I was going for was the idea that, if you choose Easy/Succeed with an explicit target, you’re more likely to hit that target (or hit it faster/cheaper/harder), but your implicit targets will get sacrificed. You risk not getting the thing you actually care about, and/or you might cease to care about it. Follow incentive gradients towards an explicit target at your own risk, etc.
Looked at that way, Invest/Succeed is an important special case of Hard/Easy: What you care about is success now, so you are happy to burn future wins for current wins. In extreme you do this at actual any exchange rate and worry about tomorrow, tomorrow. When you have a low discount rate you can afford to invest. One important split in this case is when you do hidden investment versus visible investment. Often visible investment is actually succeed rather than invest.
What made Case XI so straightforward was the realization that the other outcomes (the story everyone came away with, the promises and plans expressed, etc etc) didn’t actually impact the future at all, so hitting the explicit target was indeed what mattered. Often the times when the real direct stakes are very high, so are the stakes of who you want to be (or what you can learn, or many other things) and you have to make choices. Win the battle at any cost often does not follow from needing to win!
Pingback: Exploring Premium Mediocrity | Don't Worry About the Vase
Pingback: Expanding Premium Mediocrity | Don't Worry About the Vase
Pingback: Three replies – drossbucket
Full agreement! Except section 1. I have a visceral reaction: “Play a real guitar, nerd!”
For 300 hours of skill investment, I want skill with an actual musical instrument, even if it’s weak skill. For the record, I don’t play guitar (but I definitely don’t play rock band or guitar hero).
Yeah, I thought a little about switching examples because of that problem.
I briefly did try out a real guitar, but didn’t have the time/motivation to get anywhere. Maybe some day.
Try Rocksmith. It’s like Rock Band or Guitar Hero, with an actual guitar.
Anyway, excellent article (even though the final paragraph confused me). Would you agree that “Hard Mode” can be summarized as “don’t be a hack; don’t be a poser.”?
(To Fluffy Buffalo): I did try with a real guitar once, didn’t have the patience for it on my own. I didn’t feel like the system was especially great. Maybe one day I’ll hire someone, but likely not.
Certainly not being a poser/hack is a large part of this; a hack is in Easy Mode almost by definition. I do think that’s an incomplete summary, but I can’t complain.
This resonates a lot. Thanks for writing this. In particular for blog writing, going against current incentives for maximizing audience / engagement means less feedback to continue (at least for me). But it’s also the thing that I’d want others to do.
I think there’s something here about not wanting to go for local optimization. You’re not just eschewing short-term tradeoffs for later gain, but you’re eschewing the entire idea of tradeoffs entirely, and you move to a different worldview where it’s just about Doing The Right Thing.
I’d like for this to work. But I’m not sure Reality is as kind as to allow for it.
It’s not easy, either for writing or in general. It does get easier over time as you do it more.
My rule of thumb is that the ratio of hits-to-comments for blogs defaults to about 1000:1, so if you don’t go along the gradient you risk getting zero engagement at all and having no idea if anyone is listening. That’s pretty demoralizing, even when you’re getting paid to write; I speak from experience here. So you need to have enough slack and intrinsic motivation to keep going.
If it makes it easier, I read your blog, find it quite good, and absolutely want you to do the thing you want other people to do.
Thanks for the positive feedback and response!
The 1000:1 thing seems pessimistic (and thus probably fairly close to reality).
I’ll be mulling this over more for the next few days. The “write good stuff” vs “write clickbait” distinction was fairly salient when I was trying to write the instrumental rationality sequence.
(A friend even suggested that I repurpose my writing to a more pop sci angle, which he argued would have greater aggregate benefit due to increased views. I declined on grounds of favoring depth over breadth. And other stuff.)
I would be surprised if I stopped trying to go Hard Mode for writing in the immediate future. The justifications for selling out for greater gain (and thus invest in future impact) feel very “ugh” and aversive to me.
Pingback: Best of Don’t Worry About the Vase | Don't Worry About the Vase
Pingback: Sabbath hard and go home | Compass Rose
Pingback: Slack | Don't Worry About the Vase
Pingback: Bring Back the Sabbath | Don't Worry About the Vase
Pingback: Interesting Stuff: October 16th 2017 – Dissent
Pingback: Book Review: The Captured Economy | Don't Worry About the Vase
Pingback: Book Review: The Complacent Class | Don't Worry About the Vase
Pingback: Syllable | Radimentary
Pingback: Musings on Parenting and Slack
Pingback: An Antifragile Church for Pragmatists, Questions and Answers – Open-Sourcing All Over Space
Pingback: Privacy | Don't Worry About the Vase
Pingback: Imperfect Competition | Don't Worry About the Vase
Pingback: The Lord Giveth, and the Lord Taketh Away: Why a Life Without Limits Probably Isn’t Alive. – NOSIARCH
Pingback: Potential Ways to Fight Mazes | Don't Worry About the Vase
Pingback: Create a Full Alternative Stack | Don't Worry About the Vase
Pingback: 'Life's a Bitch': Is It? What About Karma? Let's Find Out! – NOSIARCH
Pingback: Evaluating Predictions in Hindsight | Don't Worry About the Vase
I have realized that car insurance organizations know the cars which are prone to accidents and also other risks. Additionally, these people know what style of cars are susceptible to higher risk and the higher risk they have the higher a premium rate. Understanding the very simple basics associated with car insurance just might help you choose the right form of insurance policy that could take care of your requirements in case you become involved in any accident. Many thanks sharing the actual ideas on your blog.
Pingback: The Uncertainty of Death and Taxes | Don't Worry About the Vase
Pingback: Optimistic Transhuman Agendas – L Mimilington-Mamwomsley
Pingback: 2020 Election: Prediction Markets versus Polling/Modeling Assessment and Postmortem | Don't Worry About the Vase
Can you expand on “Learn to speak the language.”? What language?
I think I primarily meant literally the language, as in whichever one people you want to deal with or read were using, instead of relying on translations, but also the metaphorical one, of figuring out how people who do the real thing talk, and how to talk to them in their own form, etc.
What was the concluding video about? (It’s no longer available.)
Frank Sinatra’s my way. I should find a new link
I thought “epistemic status” was usually used to convery how confident or certain an author is about what they’re writing about. How should we interpret your “Love the player, love the game”?
(Thank you for answering my questions, btw!)
That’s the epistemic status I encourage the reader to be in, and to convey my confidence. I was playing with the form a lot then.
Pingback: Black Sands | Placebo Parachute
Sorry about my question, but I can’t understand the point of easy mode and hard mode… So what is the point? In your opinion, doing things in easy mode lead to most of the times to little reward while hard mode lead maybe to a good reward in the future but with more stress in the present? Thank you for your patience.
I’m saying that they have both different costs and different benefits. Some things cannot be accomplished unless you’re willing to do the job right, others there’s no reason to care and doing so would be dumb. There isn’t a right or wrong answer to most of these situations, it’s more about what you care about.
Pingback: Guide to Warwick, New York | Don't Worry About the Vase