Comment Permalink

I attended the March workshop. Below is my report.

So, you might think I wouldn't get much value from a CFAR workshop because (1) I know all the instructors personally, (2) I work in the same office from which they develop and test their lessons, (3) I used to be a visiting CFAR instructor, and (4) I'm a Less Wrong veteran with 50k+ karma.

But in fact CFAR's March workshop was one of the most useful weekends I've ever experienced.

A few of the classes had little new material for me, but the classes are just meant to explain the basic tools, anyway. Most of the workshop's value comes from interacting with the teachers and other alumni to figure out how best to apply the tools to your current life situation.

First, let me give a few general notes:

The curriculum has evolved quickly. Even somebody who attended in May-July 2012 would probably get several new skills + helpful followup sessions from attending in April-July 2013. In particular, the pieces of the curriculum fit together into a coherent picture of rationality far better than they did last year.
Because the curriculum evolves quickly, some classes are more polished than others, though (AFAIK) all classes have been tested and iterated a few times.
The continuous supply of good food, snacks, and drinks helped make the workshop pleasant.
On the last day of the workshop there will be 50 exciting new things you want to do, but you should only try to implement a few at a time, with help from whichever CFAR staffer you do the regular follow-ups with.
Going forward, I will probably have nearly all MIRI staff members attend a CFAR workshop.

What concrete benefits did I get from the March 2013 workshop? There are many, and there are many promising plans that I haven't yet implemented because I'm implementing only a few at a time. Here, then, is just one of the concrete benefits I got: the Murphyjitsu tool + solved sleep problem.

Murphy's Law says that "Anything that can go wrong, will go wrong." That's not literally true, of course, but it's often the case slight bumps can derail our plans. Hence the need for contingency planning overkill.

This is one reason it's important to model your degree of agency properly. If you have strong free will intuitions, your answer to "What if something unplanned happens?" might be "Well, I'll just use my Free Will to choose to accomplish my goals anyway." But if you're modeling yourself as a kluge of mostly-Sphexish cognitive modules, then you'll say "Outside view says I can easily be derailed by small bumps, so I'd better prepare specifically for each of the most likely failure scenarios, and do some cognitive training now to make sure my brain does the right thing when one of those failure scenarios hits."

One of my top priorities after the workshop was to fix my sleep problem: I was only getting good sleep 1-2 nights a week. I had tried lots of hacks but wasn't doing enough of them consistently enough to overcome my insomnia-ish condition. In my first follow-up session, Anna helped me fix this with Murphyjitsu.

One problem was that I would stay late at the office or my girlfriend's place and then be too tired to go home where I have a good mattress and can run through my whole sleep routine. So for example: What if I'm working late and get too tired to walk home? Never stay out past 12:30am, and tell others to send me to Milvia if they see me elsewhere later than that. Make sure I always have a jacket at the office so that it's not uncomfortably cold if I want to walk home. Also, do the Fermi estimate so my brain can see that it's obviously worth it to hire a taxi to take me home and get good sleep, if that's necessary. Then, simulate in my head the situation of working late at the office and then calling a taxi ("offline training"). Etc.

After doing Murphyjitsu on my entire sleep routine, I now get good sleep almost every night.

Toward a Rationality Dojo

Eliezer's very first post on Overcoming Bias was The Martial Art of Rationality. Quick summary: in the past 50 years we've learned a lot about how to improve human judgment and decision-making (JDM). If you want to become a better "rationalist", you could develop those skills and practice them regularly — just like you would if you wanted to become a better pianist, a better chess player, or a better rugby player.

I've always wanted a Rationality Dojo, but it wasn't until I attended CFAR's March workshop that I acquired a concrete, detailed picture of what that could look like.

Based on my CFAR workshop experience, let me fill in Patri Friedman's list of important Rationality Dojo qualities:

"It is a group of people who gather in person to train specific skills." The "gather in person" part of the CFAR workshop was hugely important. I have never been so motivated to diligently learn and practice the art of rationality. Training together in person leverages good news of situationist psychology.
"While there are some theoreticians of the art, most people participate by learning it and doing it, not theorizing about it." Few workshop participants know as much psychology as Dan Keys or as much probability theory as Anna Salamon, and that's fine. CFAR staffers are around to answer questions, but participants' new powers come from repeatedly drilling Murphyjitsu, reinforcement training, "goal factoring," Fermi estimates, VoI calculations, quick odds-ratio Bayes calculations, and so on — day after day, week after week.
"Thus the main focus is on local practice groups... As a result, it is driven by the needs of the learners." CFAR's early material was driven by abstract considerations of what seemed best to teach. While that's still a factor, CFAR's current lessons have been iterated repeatedly in response to the needs of session participants, and lessons that didn't work, or didn't help people much when tested, have been discarded.
"You have to sweat, but the result is you get stronger." Doing a Fermi estimate every day is kind of a burden, but they're definitely starting to feel easier and more natural to perform. One thing CFAR has done well is to break the skills down into concrete steps (often at the 5-second level) so that it's obvious how to practice the skill repeatedly.
"You improve by learning from those better than you, competing with those at your level, and teaching those below you." CFAR workshops bring together people of varying skill levels, obviously.
"It is run by a professional... The practicants receive personal benefit from their practice, in particular from the value-added of the coach, enough to pay for talented coaches." CFAR makes it the case that there are now several people who have the full-time job of figuring out how to teach people the martial art of rationality. They have time to develop carefully crafted lessons, test them and iterate them, follow-up with people, build a large dataset of what's working and what isn't and what life outcomes follow.

A CFAR workshop isn't yet a Rationality Dojo, because it's not open every day from 10 to 8 with Fermi practice on Mon-Wed-Fri and Bayes practice on Tue-Thu-Fri, but after attending the March workshop I can finally see how such a thing could exist, if there was enough interest in at least one city to have people pay a monthly fee to develop the martial art of rationality with others, week after week.

Vaniver12y60

Even somebody who attended in May-July 2012 would probably get several new skills + helpful followup sessions from attending in April-July 2013. In particular, the pieces of the curriculum fit together into a coherent picture of rationality far better than they did last year.

I attended May 2012 and March 2013, and agree that the difference between them is perceptible. The coherence is the largest change. There were several changes in focus and presentation that made things significantly better.

For example, in May 2012 there was a session about installin... (read more)

5A1987dM12y

Not if a given participant is on a weight-loss diet and would prefer to spend the willpower needed to not eat the snack on something else, I guess. :-)

See in context

44 New applied rationality workshops (April, May, and July)

by Julia_Galef

9th Apr 2013

5 min read

44

In the early days of the Center for Applied Rationality, Anna Salamon and I had a disagreement about whether we were ready to run our first applied rationality workshops in six weeks. My inside view said "No way"; Her inside view said "Should be fine"; My outside view noted that Anna had more relevant experience than I did, and therefore cowed my inside view into grudgingly shutting up.

It turned out well. Granted, the first couple of workshops were a bit chaotic (hey, sleeping in a dogpile on the living room builds character, amiright May minicampers?). But it's clear in retrospect that we got a lot more value out of diving in than we would have from the extra time spent planning.

The "try stuff fast" habit is responsible for a lot of the techniques in our curriculum; we test out classes on each other and on volunteers, observe "Oh hey, this helps other people too" or "Oh hey, no one else thinks this is useful, turns out I'm just weird," and tweak our curriculum accordingly.

And because we cannot help going recursively meta, we've built a lot of material into our curriculum to make people better at trying things that could make them better at pursuing their goals. Quick, off-the-cuff value of information (VOI) calculations help you decide when it's worth it to spend the time, or money or risk, to try something new. Againstness helps you notice and alleviate the stress responses that can keep you from trying something, once you've noticed that you should. Comfort zone expansion is basically a "try a bunch of new things" drill.

For more details on our curriculum, check out a sample schedule. I also made a simplified map of some of our classes, so you can see how I think of them fitting into the bigger picture of rationality (click to enlarge):

To the extent that I've improved my own rationality skills over the last year, I give a lot of credit to "try stuff fast." Like many Less Wrongers I have historically been more of a "thinking about things" person than a "trying stuff fast" person; given the choice of an afternoon spent debating ignorance priors or one spent figuring out how to improve my public speaking skills, I'd pick the former every time, even though the latter would be more useful to me.

I'm partially reformed now, thanks in part to the influence of Anna, whom you'll frequently overhear saying things like "I think I'll try teaching the class as if I were Val" or "We should try a different meeting format today, it's high VOI." So now I'm much more likely to notice, "Hey, in this situation I always do X (e.g., ask for feedback later, by email), so this time let me try X-prime (e.g., ask for feedback in person on the spot) -- the cost is low and it's plausible I'll learn that I like it better than my default."

In that spirit, I recommend coming to one of our upcoming workshops in April, May or July, where you will not only be introduced to all the stuff that we've tried and found promising so far, but will also be plugged into a growing network of several hundred other thoughtful and creative people who have developed their own habits you can borrow and try (we certainly do – past participants have been the origin of some of our best material). And being surrounded by other people with similar aspirations, during the workshop and in the alumni network afterwards, is the best way I know of to keep your motivation and your discipline strong.

At $3900, it's an investment, but a low-risk one, since we have a money-back guarantee. If you don't feel like what you got out of it was worth it, we'll refund your money without hesitation or complaint.

Here are the basics:

You can apply here for any of our next three applied rationality workshops:

Friday, April 26 - Monday, April 29
Friday, May 17 - Monday, May 20
Saturday, July 20 - Tuesday, July 23

Each workshop will consist of an immersive four days at a retreat near San Francisco, training you in the art of actually using rationality. That means figuring out what your goals are, and what you can be doing to pursue them more effectively; noticing when you're acting out of habit or impulse; cultivating curiosity about the world and how it works; and learning to use both your intuitive (System 1) and analytical (System 2) thinking systems to their fullest.

We're soliciting applications not just from Less Wrongers, but from other entrepreneurs, students, teachers, scientists, engineers, activists -- anyone who is analytical, friendly, and motivated to make their own careers, personal lives, and/or societies better.

For more information on our content, check out our workshop webpage, our checklist of rationality habits, or a detailed sample schedule.

We're constantly tinkering with our curriculum (as mentioned earlier), and collecting follow-up data on what works well. So while you should be aware that our material hasn't yet been subjected to rigorous long-term studies, our alumni do tend to report that they've gotten a lot of value out of their experience. Here are a few write-ups from Less Wrongers about their CFAR workshop experience and any changes they've made as a result: toner, palladias, Qiaochu_Yuan, thejash, BrandonReinhart, ciphergoth, and a bunch of other people.

The total cost is $3900, and that includes:

Three days of classes -- Six hours of class a day, with small class sizes (4-6 people) so you get a lot of personal attention from the instructors. We rearrange those small groups several times throughout the workshop to give you a chance to get to know everyone.
One day of practice – Optional but recommended, so instructors can help you make and troubleshoot a plan to use the material going forward. (If you choose to skip this day, the total cost is $3400.)
Six weeks of personal follow-ups – Talk to our staff in one-on-one follow-ups to help you get the most value out of what you've learned.
Staying on site – We rent out lovely retreat centers (lodging and food included in the cost of the workshop) so you can get to know the instructors and other participants in the evenings, during meals, and on breaks. Evenings include everything from unconferences, to parties, to impromptu Rubix-cube lessons.
An alumni network -- You'll be included in all future CFAR alumni events, parties, online forums, and so on. We'll make every effort to connect you to alumni from other workshops with whom we think you'll hit it off or have opportunities for collaboration.

Scholarships and financial aid are available -- including for many who thought they wouldn't qualify. So if you're interested in attending, definitely apply, and mention you'd like to be considered for this. We'll set up a call to discuss.

And please don't hesitate to email me (Julia at appliedrationality dot org). CFAR staff will also be in this comment thread to field questions, and some of the alumni who frequent Less Wrong may be there as well.

Apply here (the form takes less than 10 minutes, so you should do it now rather than planning on getting to it later!).

Center for Applied Rationality (CFAR)Community Page

Personal Blog

44

Mentioned in

76A map of Bay Area memespace

New Comment

69 comments, sorted by

top scoring

Click to highlight new comments since: Today at 12:34 AM

[-]lukeprog12y390

I attended the March workshop. Below is my report.

But in fact CFAR's March workshop was one of the most useful weekends I've ever experienced.

First, let me give a few general notes:

The curriculum has evolved quickly. Even somebody who attended in May-July 2012 would probably get several new skills + helpful followup sessions from attending in April-July 2013. In particular, the pieces of the curriculum fit together into a coherent picture of rationality far better than they did last year.
Because the curriculum evolves quickly, some classes are more polished than others, though (AFAIK) all classes have been tested and iterated a few times.
The continuous supply of good food, snacks, and drinks helped make the workshop pleasant.
On the last day of the workshop there will be 50 exciting new things you want to do, but you should only try to implement a few at a time, with help from whichever CFAR staffer you do the regular follow-ups with.
Going forward, I will probably have nearly all MIRI staff members attend a CFAR workshop.

After doing Murphyjitsu on my entire sleep routine, I now get good sleep almost every night.

Toward a Rationality Dojo

I've always wanted a Rationality Dojo, but it wasn't until I attended CFAR's March workshop that I acquired a concrete, detailed picture of what that could look like.

Based on my CFAR workshop experience, let me fill in Patri Friedman's list of important Rationality Dojo qualities:

"It is a group of people who gather in person to train specific skills." The "gather in person" part of the CFAR workshop was hugely important. I have never been so motivated to diligently learn and practice the art of rationality. Training together in person leverages good news of situationist psychology.
"While there are some theoreticians of the art, most people participate by learning it and doing it, not theorizing about it." Few workshop participants know as much psychology as Dan Keys or as much probability theory as Anna Salamon, and that's fine. CFAR staffers are around to answer questions, but participants' new powers come from repeatedly drilling Murphyjitsu, reinforcement training, "goal factoring," Fermi estimates, VoI calculations, quick odds-ratio Bayes calculations, and so on — day after day, week after week.
"Thus the main focus is on local practice groups... As a result, it is driven by the needs of the learners." CFAR's early material was driven by abstract considerations of what seemed best to teach. While that's still a factor, CFAR's current lessons have been iterated repeatedly in response to the needs of session participants, and lessons that didn't work, or didn't help people much when tested, have been discarded.
"You have to sweat, but the result is you get stronger." Doing a Fermi estimate every day is kind of a burden, but they're definitely starting to feel easier and more natural to perform. One thing CFAR has done well is to break the skills down into concrete steps (often at the 5-second level) so that it's obvious how to practice the skill repeatedly.
"You improve by learning from those better than you, competing with those at your level, and teaching those below you." CFAR workshops bring together people of varying skill levels, obviously.
"It is run by a professional... The practicants receive personal benefit from their practice, in particular from the value-added of the coach, enough to pay for talented coaches." CFAR makes it the case that there are now several people who have the full-time job of figuring out how to teach people the martial art of rationality. They have time to develop carefully crafted lessons, test them and iterate them, follow-up with people, build a large dataset of what's working and what isn't and what life outcomes follow.

[-]Vaniver12y60

Even somebody who attended in May-July 2012 would probably get several new skills + helpful followup sessions from attending in April-July 2013. In particular, the pieces of the curriculum fit together into a coherent picture of rationality far better than they did last year.

For example, in May 2012 there was a session about installing and using Freemind and yEd that didn't go particularly well (if you hadn't preinstalled them, you couldn't follow along because the internet was slow), and in March 2013 there was a session about implementing Getting Things Done with Google tools and Boomerang (which also had internet troubles, but was designed as an installation tutorial you could do at a later date). I only use Freemind and yEd sparingly (the first has been almost completely replaced by Workflowy), but a capturing and reminder system is far more valuable.

[-]A1987dM12y50

The continuous supply of good food, snacks, and drinks helped make the workshop pleasant.

Not if a given participant is on a weight-loss diet and would prefer to spend the willpower needed to not eat the snack on something else, I guess. :-)

[-]lukeprog12y150

Anna's "try stuff fast" mindset was the cause of the successful May 2011 "minicamp" that largely gave us the courage to try building CFAR, and also the cause of my own lean nonprofit awakening.

[-][anonymous]12y110

Considering that this is a rationality workshop, the organisers must be commended for doing good, rational marketing by setting the price at $3900 instead of rounding it up to $4000.

[-]AnnaSalamon12y100

Also, we have more financial aid for April than for May and July; so if you've been wanting to come, you can do April (2.5 weeks from today), and money is a bit tight... do apply to April.

[-]Qiaochu_Yuan12y80

It's not quite the appointed time for my one-year retrospective on my thoughts on the January workshop, but if anyone has any questions now about the impact I've seen so far I'd be happy to answer them.

[-]Xachariah12y70

I have a couple of questions:
1) Does the $3900 cost primarily cover the cost of the workshop itself, or is it mostly used as a revenue source for CFAR to keep the doors open year-round?

2) What advantages does the weekend retreat system offer that other systems don't, specifically distance learning?

3) Are there any plans to expand into distance learning, for example into the Khan Academy, edX, Udacity model?

Everything you've mentioned sounds great, but I'm inherently skeptical of a weekend retreat format. After all, even gay cure camps 'work' and have their testimonials; how do you control for the self-help effect? Additionally, it seems odd to me to be building a program around teaching rationality in-person, when every other educational institution is trying to move in the opposite direction. I'm just curious about why you've chosen to structure the program the way it is.

[-]Qiaochu_Yuan12y130

2) What advantages does the weekend retreat system offer that other systems don't, specifically distance learning?

Most of the value that I got out of the January workshop was in having face-to-face contact with the instructors and the other students. This was enjoyable, impressed the units on me strongly, and led to many interesting, unplanned, and high-value conversations (e.g. one 5-minute conversation completely changed my attitude towards both nutrition and exercise, which I finally started paying attention to after the workshop). If you want something less squishy, networking is easier in person.

how do you control for the self-help effect?

CFAR does extensive followups with workshop participants.

Additionally, it seems odd to me to be building a program around teaching rationality in-person, when every other educational institution is trying to move in the opposite direction.

I think you're using the wrong reference class. Think about rationality as a martial art and ask yourself whether you think an online course centered around learning a martial art would be successful.

[-]Xachariah12y10

I can see the usefulness of networking. Though I don't feel like I'm in the phase of my life where I'd want to go to such lengths just to network and have excellent conversations. I could imagine that being worthwhile enough to me one day.

It's good to hear about the followups. They mentioned that in the original post, but I didn't know how extensive it was. I suppose it's unfair to demand concrete data, when dealing with such small sample sizes.

Reference class tennis, but I did learn dancing via youtube + a partner. A couple of years of dance classes were pretty useless compared to how much I learned after a month of study online.

[-]Eliezer Yudkowsky12y110

I bet that partner thing was kind of important.

[-]Xachariah12y20

Are you suggesting that rationality takes the same level of one-on-one contact that dancing does?

I'm sure it wouldn't be too much trouble to go out and find rationality partners, if that's what it took. I'd still need the curriculum though.

[-]Viliam_Bur12y20

Talking with a rational person who disagrees with you can be a great experience, if both of you prefer finding the truth above winning the discussion.

(I am not saying that the minicamps are about this. Just answering your question, out of context.)

[-]DaFranker12y20

I bet that "kind of important" thing was kind of an understatement.

[-]palladias12y70

Re (1): The $3900 covers the cost of the workshop (the earliest workshops were offered at a loss to start testing the material). It covers cost of venue, outreach to find participants, and time spent on curriculum development and research to make sure the material offered is useful and clear.

[-]Kevin12y40

1) There isn't that much of a pragmatic difference between whether the money is going to the workshop cost or to keep the CFAR doors open year round. CFAR needs to keep the doors open year round in order to continue to be able to host workshops and keep developing and refining curriculum. The venues and food are always great, but not $1000/night nice.

2) There's something about the weekend retreat format that allows for a really strongly transformative experience. Your time at a CFAR workshop somehow feels more significant than your time in everyday life, like each moment is just pulsing with more of the very fiber that gives reality its existence. This also results in some optimization for remembering self -- the moments at the CFAR workshop feel like something that will be remembered and deconstructed for far longer than the experience itself lasts. And there is a neat change towards increased emotional openness that happens over the course of spending such an intense weekend with new friends. Also, the kinds of people that attend $3900 weekend workshops can make for incredibly high value networking.

3) I expect it will happen eventually, sure.

[-]FiftyTwo12y70

Are there any plans to run them in other countries? Even with the posibility of scholarships international travel isn't cheap.

[-]AnnaSalamon12y180

Alas, we're unlikely to do this in the next 12 months.

Although: if you or someone else based in a foreign country wants to: (a) do all necessary research to figure out whether we're legally allowed to run a workshop in that country (given its visa requirements, the fact that we aren't citizens, etc.); and (b) do the leg work involved in finding a retreat site, and in finding a bunch of local-ish participants who wish to attend... we would then strongly consider it. Do let me know if you're interested.

[-]Kaj_Sotala12y70

Not directly related to the post itself, but I thought that I'd drop the link to this recent paper here so that CFAR folks see if it they haven't already: http://www.sciencedirect.com/science/article/pii/S1364661313000405

Although human thinking is often biased, some individuals are less susceptible to biases than others. These individual differences have been at the forefront of thinking research for more than a decade. We organize the literature in three key accounts (storage, monitoring, and inhibition failure) and propose that a critical but overlooked question concerns the time point at which individual variance arises: do biased and unbiased reasoners take different paths early on in the reasoning process or is the observed variance late to arise? We discuss how this focus on the ‘whens’ suggests that individual differences in thinking biases are less profound than traditionally assumed, in the sense that they might typically arise at a later stage of the reasoning process.

[-]lukeprog12y100

PDF.

[-]Nornagest12y60

Since the potential for a "those who can't, teach" thing seems to be a sticking point here, I wonder how hard and/or legally complicated it'd be to set up a controlled experiment. Keep the $3900 price point for classes that are guaranteed to represent CFAR's best effort at providing rationality skills, but also offer a limited number of slots at a lower or zero price point that might be a best effort or might just contain impressive-sounding platitudes of the sort that you hear in your average motivation seminar. Follow up with students after six months or a year and see what the deltas on achievement are.

[-]Richard_Kennaway12y70

Here's an idea for a smaller-scaled test. Begin a lecture to various audiences (CFAR camps, TED, etc.) by spouting a lot of impressive-sounding platitudes and see how fast they pick up on it.

[-]Nornagest12y40

That might show that CFAR's audience, or some of its members, wouldn't be gulled by a placebo self-help program, but it wouldn't show that CFAR's program is effective -- it might consist of more sophisticated platitudes, or teach skills that're only useful in unrealistically simple situations, or be impractical in other ways. If you want to know about effectiveness, you at some point need to compare outcomes.

A cheaper test might involve collecting outcome information from a sample of CFAR students and then finding similar samples from other programs and following up with them, but that would be susceptible to a bunch of demographic confounders.

[-]private_messaging12y50

Yes. You'd also need a fairly large sample size.

The issue isn't only that those who can't, teach. What if use of some math-inspired insights is actually very difficult? (Who'd offer 4 day classes in that then?) .

[-]MugaSofer12y-20

Not sure how much that would show, but sounds worth trying anyway.

[-]private_messaging12y10

Well, it wouldn't be blind this way. The attendees of the usual platitudes-filled self help workshop absolutely swear by the gains. There may actually be gains - some people really just need to be reaffirmed via platitudes.

[-]Nornagest12y10

I'd expect to see some gains in both forks of the experiment, for exactly those reasons. The question is whether or not we'd see more gains from a proper rationality course than from a motivation seminar dressed up as one: if so, that's good evidence that CFAR is teaching something worthwhile. If not, well, that would also be nice to know.

Blinding it would be hard, since we're dealing with smart people who'd have some reason to be skeptical of the course contents and especially since many of them would have read LW and have some idea of what a rationality program should look like, but I think it should be doable.

[-]private_messaging12y60

I dunno what an actual rationality program should look like, but in general I'd say it would have much more emphasis on time. Smart but inefficient people I know are already spending too much time in analysis paralysis(i.e. basically doing what's on this graph out there) rather than doing things that generally get them closer to multitude of goals under a multitude of possible assumptions about the world.

edit: one curious thing about the world is the enormous size of solution space. When you are not sure what you'll want, when you're not sure of some facts, you can restrict the search space to what furthers either goal under either assumption of the facts - you'll still miss all the really good solutions even in that restricted space due to inability to search all of it, and this restriction decreases expected value of a solution you may find by a negligible amount in a lot of the cases. This is dramatically different from most textbook examples on the subject. Just think about it. There's probably a tweet-sized text that you could, in theory, come up with which constitutes an insight for an invention on which you can make millions dollars. You can search for this, or you can, figuratively speaking, ponder which can of juice at the supermarket is cheaper per litre or try to quantify how much you like either juice and multiply that with price.

[-]Kawoomba12y50

Yea, there is often an emphasis on how an ideal unconstrained agent (with a resource-unlimited AIXI like superintelligence in mind) would reason, or how biases should be avoided, with the assumption that a resource-limited agent should just try to approximate what a (practically) resource-unlimited agent should do as closely as it can given its resources.

That, however, is a bad assumption to make. For every limitation-class, an entirely different strategy and different heuristics may be optimal.

As an example, when you try to program your old calculator to play speed chess as well as possible, you should not try to take the best available chess program. There are biases you should adopt to get the best result, given the limitations. I was once in a grad AI class in which we needed to best a given computer opponent in an obscure game, with severe computational resource limitations. The best program was a seemingly bad hack of in generality wrong heuristics, but it turned out to use the resources in the most optimal way. It bested much "better" programs that would've beaten it if given more time.

Take teaching the Bayes equation as if that's what you'd actually use. Sure, some general ideas (consider the prior, always update on your observations) are useful, but the equation itself? Noone at CFAR walks into a supermarket and then continuously inputs actual numbers into Bayes equations in their head.

[-]palladias12y50

Take teaching the Bayes equation as if that's what you'd actually use. Sure, some general ideas (consider the prior, always update on your observations) are useful, but the equation itself? Noone at CFAR walks into a supermarket and then continuously inputs actual numbers into Bayes equations in their head.

Agreed! I teach the Bayes class at workshops, and it's not a math drill class. It's on how to get the habit of paying attention to the components of Bayes theorem in everyday life. For example, we usually ask just ask "Would I be likely to see Y if X were true?" and skip the question "Would I be likely to see Y if X were not true?" So we practice ways to trigger this thought so you don't get tricked by base rates or other pitfalls.

Concrete example: Someone you're interviewing for a job flubs one question and your first thought is that your shouldn't hire them, because people who aren't qualified flub questions. But pause and ask how often you expect qualified people to miss one question in an hour long interview. Your answer will vary based on the kinds of questions your asking, but you may be treated the evidence as a stronger signal than it is.

[-]private_messaging12y10

Concrete example: Someone you're interviewing for a job flubs one question and your first thought is that your shouldn't hire them, because people who aren't qualified flub questions. But pause and ask how often you expect qualified people to miss one question in an hour long interview. Your answer will vary based on the kinds of questions your asking, but you may be treated the evidence as a stronger signal than it is.

Or maybe you treated it as a weaker signal than it is. This is a strawman anyway, people who never in their life heard of Bayes do compare it to their hypothetical idea of how a competent person would do on their exam, and soon thereafter, to their actual knowledge of how a competent person does, remedying all sorts of miscalibrations.

If anything, in practice the actual problem with interviews is generally that incompetents get through, because incompetents are being interviewed so much more than anyone else. Diligent ability to never flunk anything (conscientiousness) is, at least, something very useful in workplace that you can't fake by preparing specifically for interviews.

Then there's also this enormous utility disparity between the minor dis-utility of perhaps running the interviews for a little longer and ending up hiring the best when no one passes, and major dis-utility of hiring an incompetent.

It's not really an advice, but I can see how its likeable - there's people who didn't get hired because they flunked "maybe one question", and these folks will get a fix of their endorphins when they rationalize it as the HR person being irrational.

[-]PrawnOfFate12y10

A good test will include a few questions almost no-one can answer. That avoids the problem of having more than one 10/10 score.

[-]Kindly12y40

More generally, a very good test will result in a uniform distribution of scores (rather than a bell curve), maximizing the information content of the score.

[-]private_messaging12y30

Well, even bigger issue with Bayes is how easy it is to get it wrong on graphs in general (which contain loops). Worse than that, what we actually have is uncertain topology. This all should make rational updates much smaller and compartmentalized-looking than some naive idea of 'updating' beliefs from one argument, then from the other, and so on.

There's also a lot of very advanced math related to specifically estimation. E.g. an expected utility would be a huge sum, vast majority of terms in which you do not know. When deciding on some binary action, you have those two sums on both sides of comparison, and you need to estimate the sign of the difference most accurately (which is dramatically not same as summing all available terms), then you need to quantify the expected inaccuracy in your estimate of the sign, and adjust for that. Simply put, its complicated and people who have good working understanding of such concerns can write important textbooks, software, papers, etc. (Which make a lot of difference to the world, as well as make any spin off 'workshops' credible). Whereas people who are very far from having any understanding of such can do things like estimating 8 lives saved per dollar donated.

[-]HungryHippo12y30

There's probably a tweet-sized text that you could, in theory, come up with which constitutes an insight for an invention on which you can make millions dollars.

Thank you for this. You managed to translate "the world is full of possibilities" into something that hits right in the gut.

[-]A1987dM12y50

They offer a refund if you attended a workshop and didn't like it. If you didn't... what are you talking about? :-)

[-]V_V12y-30

The point is not whether you like the workshop, it's whether it teaches you something worth its price.

[-]A1987dM12y30

I'd think that, among the people who would consider attending such a workshop in the first place at least, the two are highly correlated. (Also, I meant “didn't like” in a general sense, such as ‘regret’.)

[-]V_V12y-40

Why?

[-]A1987dM12y20

Well.. I'd guess people go there to learn stuff, so they like it if they do learn stuff.

OTOH, they might at first be wrong about how useful the stuff they learn will turn out to be. For how long after the workshop will they accept requests for refunds?

[-]AnnaSalamon12y130

One year.

[-]V_V12y-40

OTOH, they might at first be wrong about how useful the stuff they learn will turn out to be.

Indeed.

They might be also permently wrong about how useful the stuff they learn will turn out to be:
For instance they might attribute failed outcomes to failure to apply what they learned rather than to an inherent flaw of what they learned.

Or they may ignore/downplay failures and exaggerate successes: that's good old confirmation bias and sunk cost fallacy (there are some costs, such as time spent and travel expenses, which wouldn't be covered by a refund).

Admitting that they invested time and money on something that wasn't worth its price, and that they didn't find out immediately, would hurt their self-image. IIRC, even the victims of outright illegal scams often don't report them to the authorities, even after they realize that they have been scammed. CFAR workshops most likely aren't illegal scams, thus I expect that the resistence to seeking reparations would be even greater.

There might be further reasons not to ask for a refund: for instance people might want to attend to these workshops to associate themselves with high-status members of the so-called "rationalist" community. They might fear that asking for a refund might be seen as defection.

[-]Viliam_Bur12y00

CFAR workshops most likely aren't illegal scams

And you, sir, most likely are a troll.

[+]V_V12y-130

[+]private_messaging12y-80

[-]Shmi12y50

At $3900, it's an investment, but a low-risk one, since we have a money-back guarantee. If you don't feel like what you got out of it was worth it, we'll refund your money without hesitation or complaint.

This is still not low-risk. I would hesitate to ask for a refund even if an event like this was below my expectations, as long as it's not a total flop or a con, which it surely isn't. Low-risk (for the participant) would be dividing the camp into billable events with a price tag on each, and refunding a portion of the price of each event based on the post-event evaluations. This is probably unworkable in practice, but at least it would not be misleading. On the other hand, "full refund no questions asked" is a useful marketing strategy, if a bit dark-artsy.

[-]Julia_Galef12y210

If it makes you feel less hesitant, we've given refunds twice. One person at a workshop last year who said he'd expected polish and suits, and another who said he enjoyed it but wasn't sure it was going to help enough with his current life situation to be worth it.

[-]Shmi12y50

Oh, the refund clause would not have mattered to me personally, for the reasons outlined (I know I would never ask for one, no the least because I would thoroughly enjoy the event). I would dearly love to attend, but for the reasons I am not willing to discuss here it would not be a rational decision for me. My comment was just an observation that your claim of low risk is not really accurate, except for a rare person who has a certain mindset.

[-]ThoughtSpeed12y40

At what point do you guys estimate CFAR will scale such that economically disadvantaged individuals such as myself will be able to afford a retreat? In the next few years will there be more of a focus on making money off of increased demand from business, heavier-pocketbook type individuals, or lowering costs for hungry student types?

I would love nothing more than to go, if only it was cheaper.

[-]JGWeissman12y100

OP says:

Scholarships and financial aid are available in some circumstances. So if you're interested in attending, definitely apply, and mention you'd like to be considered for this. We'll set up a call to discuss.

[-]Viliam_Bur12y80

Data point: me and my girlfriend did get a scholarship last year. (But if we were literally hungry students, even that probably would not be enough.) For us, the most expensive thing was actually travelling to USA. So for people like us it would be most helpful if CFAR could expand to other continents. But I completely understand that this is not the top priority of CFAR today. It could be a higher priority for people living far from the Bay Area. So if your situation is similar, starting a rationalist group in your community could be helpful. Just show people around you LW and HPMoR and see what happens. I hope that within a year or two it will be possible to have a CFAR (or CFAR-like) workshop in Europe too.

PS: I know some people value their internet anonymity, but if you disclose where you live, you increase your chances of meeting other rationalists personally, which is a great experience. With a small group, you could already start doing some rationality exercises described on LW.

[-]Dusk12y20

It is great to have the dates. I had been hoping for an event later in the year. What are the odds that other mini-camps will be run between September and December?

[-]JMiller12y10

Hi, the "apply here" link is not working for me.

Thanks!

[-]Julia_Galef12y00

Fixed! Thanks, I apparently didn't understand how links worked in this system.

[-][anonymous]12y00

Here are a few write-ups

First link ("toner") is broken (missing domain name). Please don't reply to this comment: I plan to delete it once the link is fixed.

[This comment is no longer endorsed by its author]Reply

[-]dspeyer12y00

I also made a simplified map of some of our classes, so you can see how I think of them fitting into the bigger picture of rationality (click to enlarge):

I've clicked everything I can think of and it's not enlarging.

[-]Julia_Galef12y10

Fixed now, sorry!

Moderation Log