Matthew Barnett's Shortform

Matthew Barnett

This is a special post for quick takes by Matthew Barnett. Only they can create top-level comments. Comments here also appear on the Quick Takes page and All Posts page.

I generally agree with the heuristic that we should "live on the mainline", meaning that we should mostly plan for events which capture the dominant share of our probability. This heuristic causes me to have a tendency to do some of the following things

Work on projects that I think have a medium-to-high chance of succeeding and quickly abandon things that seem like they are failing.
Plan my career trajectory based on where I think I can plausibly maximize my long term values.
Study subjects only if I think that I will need to understand them at some point in order to grasp an important concept. See more details here.
Avoid doing work that leverages small probabilities of exceptionally bad outcomes. For example, I don't focus my studying on worst-case AI safety risk (although I do think that analyzing worst-case failure modes is useful from the standpoint of a security mindset).

I see a few problems with this heuristic, however, and I'm not sure quite how to resolve them. More specifically, I tend to float freely between different projects because I am quick to abandon things if I feel like they aren't working out (compare this to the mindset that some game developers have when they realize their latest game idea isn't very good).

One case where this shows up is when I change my beliefs about where the most effective ways to spend my time as far as long-term future scenarios are concerned. I will sometimes read an argument about how some line of inquiry is promising and for an entire day believe that this would be a good thing to work on, only for the next day to bring another argument.

And things like my AI timeline predictions vary erratically, much more than I expect most people's: I sometimes wake up and think that AI might be just 10 years away and other days I wake up and wonder if most of this stuff is more like a century away.

This general behavior makes me into someone who doesn't stay consistent on what I try to do. My life therefore resembles a battle between two competing heuristics: on one side there's the heuristic of planning for the mainline, and on the other there's the heuristic of committing to things even if they aren't panning out. I am unsure of the best way to resolve this conflict.

Mathematically, it seems like you should just give your heuristic the better data you already consciously have: If your untrustworthy senses say you aren't on the mainline, the correct move isn't necessarily to believe them, but rather to decide to put effort into figuring it out, because it's important.

It's clear how your heuristic would evolve. To embrace it correctly, you should make sure that your entire life lives in the mainline. If there's a game with negative expected value, where the worst outcome has chance 10%, and you play it 20 times, that's s

... (read more)

2Trinley Goldenberg7y

There are two big issues with the "living in the mainline" strategy: 1. Most of the highest EV activities are those that have low chance of success but big rewards. I suspect much of your volatile behavior is bouncing between chasing opportunities you see as high value, and then realizing you're not on the mainline and correcting, then realizing there are higher EV opportunities and correcting again. 2. Strategies that work well on the mainline often fail spectacularly in the face of black swans. So they have a high probability of working but also very negative EV in unlikely situations (which you ignore if you're only thinking about the mainline). Two alternatives to the "living on the mainline" heuristic: 1. The Anti-fragility heuristic: * Use the barbell strategy, to split your activities between surefire wins with low upsides and certainty, and risky moonshots with low downsides but lots of uncertainty around upsides. * Notice the reasons that things fail, and make them robust to that class of failure in the future. * Try lots of things, and stick with the ones that work over time. 2. The Effectuation Heuristic: * Go into areas where you have unfair advantages. * Spread your downside risk to people or organizations who can handle it. * In generally, work to CREATE the mainline where you have an unfair advantage and high upside. You might get some mileage out of reading the effectuation and anti-fragility sections of this post.

6Raemon7y

Some random thoughts: * Startups and pivots. Startups require lots of commitment even when things feel like they're collapsing – only by perservering through those times can you possibly make it. Still, startups are willing to pivot – take their existing infrastructure but change key strategic approaches. * Escalating commitment. Early on (in most domains), you should pick shorter term projects, because the focus is on learning. Code a website in a week. Code another website in 2 months. Don't stress too much on multi-year plans until you're reasonably confident you sorta know what you're doing. (Relatedly, relationships: early on it makes sense to date a lot to get some sense of who/what you're looking for in a romantic partner. But eventually, a lot of the good stuff comes when you actually commit to longterm relationships that are capable of weathering periods of strife and doubt) * Alternately: Givewell (or maybe OpenPhil?) did mixtures of shallow dives, deep dives and medium dives into cause areas because they learned different sorts of things from each kind of research. * Commitment mindset. Sort of how Nate Soares recommends separating the feeling of conviction from the epistemic belief of high-success... you can separate "I'm going to stick with this project for a year or two because it's likely to work" from "I'm going to stick to this project for a year or two because sticking to projects for a year or two is how you learn how projects work on the 1-2 year timescale, including the part where you shift gears and learn from mistakes and become more robust about them.

Matthew Barnett's Shortform

7

Ω 2

Matthew Barnett's Shortform

7

Ω 2