Appreciate you writing up your thoughts!
When you compare their reported work/career satisfaction improvements to the clinical effect size of antidepressants (typically around d = 0.3 over placebo), the results are impressive
I don't think this is a fair comparison. Usually clinical trials will involved some sort of control group. If you're referring to the 2015 longitudinal study, then there was no placebo/control group. You can't be sure whether the retreat itself had any influence on the measured improvement. Attending the retreat might just select for the kinds of people who are looking to improve in their career/work since it (presumably) selects for those interested in 'self improvement' (scare-quotes since this makes it sound like self-help literature, which I don't think its accurate).
That is a really good point, and it helps me think much more clearly about the comparison I made.
You're completely right about the lack of a control group. If we assume that tracking any similar cohort (matched for age, gender, education, and an active interest in self-improvement) would naturally show similar upward trends over a year, then attributing the effect entirely to the retreat is definitely flawed.
When I wrote that, I was operating under the baseline assumption that without a specific treatment or change in circumstance, life satisfaction and those other metrics would remain relatively static. However, I haven't actually read the literature on the natural trajectory for this specific demographic—it's entirely possible that a similar cohort would see an organic increase anyway simply because they are highly motivated people looking to improve. I don't have enough confidence to make a strong claim either way.
A much better study design would use an active control group, for example, comparing CFAR attendees to people attending Less.online meetups or similar events. That would help us figure out if the measured effects actually come from the CFAR curriculum, or if they just stem from the baseline benefits of aggregating rationalists together and having a fun gathering.
Why did I write this?
There is surprisingly little information online about what actually happens at a Center for Applied Rationality (CFAR) workshop. For the only organization that teaches tools for rationalists in real life (AFAIK), the actual experience of the workshop has very few mentions [1]. (Though recently, Anna Salamon has been making more posts [2]).
I wanted to write something short and concrete to record my experiences. If there is interest, I can provide more details and answer questions.
Why did I go?
The pitch for CFAR usually goes something like this:
If you run the Expected Value (EV) calculation on "having better thinking tools for the rest of your life," the numbers get silly very quickly. You can easily conclude that you should at least investigate. So I did.
Unlike many other corporate retreats or workshops, there is some evidence backing up this claim. A 2015 longitudinal survey [3] followed up on CFAR participants (n=135) by comparing their answers pre-workshop and post-workshop across four areas: well-being, personality, behaviors, and productivity.
They found significant effects in many areas. When you compare their reported work/career satisfaction improvements to the clinical effect size of antidepressants (typically around d = 0.3 over placebo), the results are impressive:
I had the free time, the EV calculations worked out, and I was interested in talking to more rationalist folks. So I went.
So how was my experience?
1. Does it actually teach the techniques well?
Is it better than just reading the handbook [4] by myself or with my local rationality group?
Short answer: Yes.
The format (6-10 students, 1 instructor) works well. The sessions I enjoyed most started with ~20 minutes of the instructor giving practical examples, followed by ~40 minutes of paired practice with a worksheet. Students are encouraged to ask questions during both sections, and the small group size generates useful positive and negative examples of applying the technique.
I really enjoyed the "Finding Cruxes" workshops. I’m familiar with the theory, but actually having a trusted peer sitting across from you, both trying to notice the crux while keeping track of the argument itself, is much more practically useful than reading a blog post.
However, there is high variance in the classes. Some of the theory-heavy or ideology-heavy classes went over my head (though I noticed some of the more practiced rationalists enjoyed them, so it’s potentially an experience gap problem). Other classes helped reframe my problems, leading to some "wow, I never thought of it like that" moments.
2. Is it fun?
Yes. I enjoyed it more than the counterfactual use of my time. We joked that we were a group of "social autists" (with a potentially diagnosable rate of ~40%), so the social norms were explicitly designed for people like us. It is simply fun to hang out with people who share your inferential framework.
3. Is it useful?
Anna Salamon suggests that a lot of the time, a technique is supposed to feel like a "mental trick." While not as rigorous as mathematical equations, it is helpful for reframing a problem in an easier way.
For example, in the Question Substitution class, I realized that I judge other people’s experience of happiness by modeling it on my own mind. That is a simple, obvious error. But until I consciously thought about explicitly swapping the question and went through the worksheet, I hadn't noticed I was doing it.
Personally, I solved some of my problems during the retreat. I suspect the tools taught are supposed to help with epistemics and treating your emotions like unconscious signals rather than white noise, which is something rationalist types probably should do more of.
What Actually Happened? (Logistics)
For those unaware, CFAR has been running on and off since 2012. After a hiatus, they "renewed" operations in 2025 with workshops in Austin and SF[5]. I went to the Austin one.
The structure was a 4-day retreat on a ranch:
We also had 2-3 hours of break for lunch and dinner in between, so it was a very comfortable but packed schedule.
The 20-Hour Win
Was any of this useful? For me personally?
Yes. I saved about 20 hours (minimum) of work on my current research problems just by talking through them with workshop instructors during the Questing [6] activity.
Though this was more along the lines of professional advice, could I have gotten this elsewhere? Probably. But I don't think I would have brought this up without a similar environment to CFAR, with like-minded peers and this level of vulnerability.
Notably, the classes are high variance. A CFAR instructor said something to the effect of:
I did feel like I came away with at least 2 out of the last 4 skills.
That said, I took a day of flights from Australia to get there. So, depending on how you value your time, the net EV might still be negative. /shrug.
What’s Next?
For me:
For rationality workshops in general:
I noticed that a lot of the value comes from the instructors and experienced participants, many of whom are US-based. CFAR really is "Community + Practice."
I expect it will be very difficult to replicate this if you aren’t in a rationality hub like the Bay Area or Austin. A small local practice group going through the materials might get you 20% of the "goodness" of CFAR, but you’ll miss the emotional connection and vulnerability part that comes from the immersive retreat.
Acknowledgements
I want to give a huge shoutout to Wendy and the logistics team. Great work handling the storm and providing a comfortable space for everyone. It is important and meaningful work! Also, a shoutout to the friendly Austin folks; I really appreciate the hospitality!
Also Big thanks to the nice folks at CFAR for the experience. Any mistakes in explaining them are 100% mine.
References
[1] LessWrong Tag: Center for Applied Rationality
[2] User: Anna Salamon
[3] CFAR 2015 Longitudinal Study
[4] CFAR Handbook Introduction
[5] CFAR Update and New Workshops (2025)
[6] In Question we were paired up with a partner and take 15 min turns of just watching the other partner try to do something. Its suprisingly useful the simple idea of borrowing 15 mins of a trusted peer's time.
[7] CFAR Takeaways (Andrew Critch)