Locally optimal: when problems are solutions
There’s no way that chronic depression, self-loathing, poor agency, or muscle tension could be optimal… right?
Jake felt horrible every time he interacted with other people. He didn’t know what to do about it.
So he had an incentive to avoid interacting with others, and no way of making it feel safe. Guess what happened next!
He became depressed.
Depression certainly wasn’t globally optimal, but it was locally optimal.
What he thought was a “problem” was actually a solution — to stop feeling horrible when interacting with others, he just *stopped* interacting with people!
This changed later once Jake unlearned the insecurities that made social interaction feel unsafe in the first place.
Because there wasn’t a need for the “depression strategy” anymore, his symptoms evaporated.
I know because Jake was me.
Locally optimal strategies are common
It wasn’t just my depression and anxiety that were locally optimal strategies, but also my conflict avoidance, emotional numbness, eye contact, boundaries, empathy, and neck pain. I didn’t realize it until I checked, but I felt that bad things would happen if I resolved each issue. So the issue had incentives. None of these issues improved until I addressed their incentives.
Take my chronic neck pain. For 3½ years I had chronic tension that was sometimes so bad I couldn’t turn my head. The first approach I tried was physical therapy and stretching. This approach treated the tension as simply suboptimal, as if it was only accidentally tense. And while this could be the case for other people, my own neck pain barely improved. Years later, I used coaching to feel into what would happen if my neck relaxed, discovered the tension was helping me avoid interpersonal conflicts in an unexpected way. I integrated this (found a way to achieve the same goal while staying relaxed), and have had ~90% less neck pain ever since.
After resolving many of my issues by feeling the incentives, I wondered whether it could help others. To validate this hypothesis, I began offering pay-on-results coaching (bounties = data collection). It’s helped people resolve lifelong issues resistant to 8 years of talk therapy, 2 years of IFS, hundreds of hours of meditation, CBT, psychedelics, etc.
As it turned out, previous approaches had never treated their issue as locally optimal:
“How are you feeling this week? Let’s talk about that.” “Tell me about your childhood.” “Meditate more.” “Trip acid.” (Attempts were not goal-oriented.)
“Your insecurities were useful at some point in the past, but now they’re just old strategies that aren’t serving you anymore.” (Treating the issue as suboptimal, not presently locally optimal.)
“Meditate so you can be with those feelings.” (Maybe avoiding feelings itself is a strategy!)
Meanwhile in the locally optimal approach, I show them how to look for bad feelings if their issues are actually resolved:
People with social anxiety or self-loathing may want to feel confident, but are afraid of having more interpersonal conflicts.
For one client, lack of agency looked like a way to avoid perceptions of failure.
For another, lack of romantic confidence looked like a way to avoid successful relationships. This was bad because he didn’t feel deserving.
These incentives need to be integrated, not ignored.
When an issue has lasted for years, it’s likely locally optimal. This is because: if there are zero downsides to improving an issue, then why hasn’t it improved already? (In contrast, new issues are more likely to be simple skill issues without hidden incentives.)
How can I use this?
You will not be able to read this post and coach yourself. I generally recommend against self-coaching.
Instead, use this idea to evaluate who to hire once you have the funds. I would find a skillful coach or therapist who… you vibe with, believes they can help your issue, uses some kind of somatic practice (e.g., Gendlin’s Focusing), and also believes that chronic issues are usually useful strategies, right now in the present day. Many coaches and therapists say things like “These issues were useful at some point in the past, but they don’t serve you anymore” which completely fails this test.
Remember: Many coaches and therapists don’t help resolve issues. The best therapists are 10x more effective than average.
Thanks to Brian Toomey (Harvard Meditation Research Program, author on emotional change and memory reconsolidation), Kaj Sotala (Multiagent Models of Mind), Stag Lynn, Ethan Kuntz (QRI), and Anna Salamon.
Here’s a post that applies this idea:
