amp vs leverage

Amp is imaginary.

How she came about is interesting but not important here. You can think of her as like one of those parts of my personality that I imagine chatting to (and sometimes write those chats here), but with a slightly more permanent label. I don’t invoke her that often now, but she can be useful as a way of packaging up advice that I want to give myself.

Leverage Research is a facility somewhere out in the Bay Area, made up largely of people who identify as effective altruists. They are very secretive – they’ve run some very visible projects including the first few EA conferences and “THINK” – the first organized attempt to bootstrap a global network of local EA meetings. But what they do the rest of the time is anyone’s guess. Their boss is very charismatic and apparently has a good knack of getting people to fulfill their potential. My hypothesis is that this is a big reason people put up with Leverage.

They like flowcharts.

They also like something called introspection. Their website is still fairly minimalist, but there’s now a blog section and it’s talking about this. It means thinking about your own mind, with a view towards not yielding answers that are ridiculous. (That’s the hard part of course. It’s easy to think about yourself but hard not to jump to the wrong conclusions).

Out of this emerged something called Folding.

I won’t try and describe the concept or any of the theory that motivates it, because I wouldn’t do it justice and it might all be nonsense anyway. You can read about it yourself.

But last night I was thinking in a vaguely Folding-like way. I was imagining Amp (who does not correspond to any kind of Folding concept or technique) and how I would ask her how to start optimizing my life again.

One of the concepts from Folding – not what it’s primarily about but a necessary prerequisite – is Level 2 concepts. These are basically your mind’s internal language of ideas which haven’t yet been formed into words – or never will. Forming them into words isn’t the goal necessarily – you can work with them directly and can sometimes achieve things with that that mere language is too crude and clumsy for.

I was approaching the question of “how should I go back to optimizing myself”, not intending to do it – because I was retired to bed for the night and there wasn’t much I really could do – but more intending to think about it a bit. A level 2 concept appeared: in words, it would correspond to something like “When you intend to optimize yourself, ugh you are referring to [..some process..]. [sense that something else is possible]”

That of course can be expressed in words, and sometimes level 2 thoughts have sufficient structure that it’s worth doing that. In words though, it has several subclauses and takes a few seconds to process – in level 2 it’s just a single thing (or maybe two things that happen together – the yucky concept of what I mean by optimization and a separate shadowy sense that something else is out there, without having a clear handle on what it might be).

Unfortunately, though, I don’t yet know what my mind was trying to tell me. Generally our linguistic thoughts are somewhat robust to interrogation – if something gets as far as words, in my mind at least, then I can usually somehow justify it, expand upon it or at least explain what the thought might have been caused by. Once I start pulling these little artifacts directly out of the mental murk, however, what they’re actually talking about is anyone’s guess.

At least, to begin with.

There was one other interesting thing that came up in this session, before I gave up on it and fell asleep, and it’s the subject of this post. Amp, my imaginary thought persona, had been helping me through this folding exercise and I randomly stumbled on the idea of imagining Amp working at Leverage Research. This created a momentary feeling of dissonance and I wasn’t immediately sure why.

The obvious explanation – that Leverage Research only hires people who actually exist – didn’t really resonate. What came up instead was this idea of fake optimization.

Whatever it is that the people at Leverage are doing, they work really really hard at it. They’ve come up with a bunch of techniques for how to squeeze more ours out of the day – everything from messing with sleep patterns to cutting entire sections of your life out that you don’t feel you need any more. And they employ those techniques themselves. But… that’s difficult to do, you really need to know what you’re doing with that.

What I do instead, when I get really into something like trying to understand this MIRI paper, is that I go all nuts for it, putting in any available hour. Any general tasks for maintaining the rest of my life get put on hold – and it’s not like they were getting much attention in the first place. The “Amp” side of my mind is clearly worried about this: her advice would be to tackle what’s actually the most important problem in my life first, and it’s not that I don’t know enough math. Right now she’d be telling me that I need to set some boundary around how much I work on this math thing, and let the people in the forum know how much work they can expect from me.

Even though I wouldn’t expect actual members of Leverage to disagree with that, somehow what they embody in my mind comes into conflict with it. (And of course the feeling of dissonance somehow encompassed more than that: but this was a concrete thing I could get a handle on).

And so this thing about fake optimization – working really really hard at something that seems sort of important or interesting in a far-mode sense, at the expense of the boring immediately-needing-to-be-done stuff – came up. But the thing is, it’s not like I didn’t know about it before. I was already aware that I behave in the way that I describe, and if I consciously set down to prioritize what needs to be done next then it’ll come out as something sensible.

And yet… another of those feelings of dissonance, just as I write that.

That idea of “sensible”… That’s another of those “ugh you mean [..some concept..]” things. There’s a feeling of being a multi-agent system again: the part of me that’s “sensible” is the part that comes to the foreground when I write to-do lists, there’s a sense of another part that stops it from coming to the foreground in the first place because it knows what it’ll do.

Neither of those are Amp by the way. Ignore Amp right now.

There’s a sense in which these two actors take on the roles of the official self and the shadow self, the one that represents what you want to want and the one that represents what you really want and are scared to admit. Or however you wanna think of them.

The shadowy one is sort of dominant right now. After all, I’m not actually writing that to-do list that would contain at the top: stop writing stupid blog posts and start doing one of those useful things you’ve been putting off, look you’re almost in the mood to anyway and really now’s your chance otherwise you know what’ll happen.

What happens if I gently intend to write a to-do list? Obviously I mean fake-intend, not the decision-theoretic sense of intending where there has to be a serious danger that I’ll actually end up doing it. But just gently poking my thoughts in that direction and seeing which other thoughts come up.

I mean there was a warning that “you know you can’t win”. Writing to-dos is a short term win: some of them will get done (or when it’s really bad, when I set off to do one of them but end up at 3am or 4am having done nothing and hating myself), and then I’ll be back to where I was. It’s constantly uphill. There’s a sense in which I could win by just constantly going uphill, and at least I’d get more done than I am now, but that warning voice is again warning that there’s no outside-view reason that’s going to happen any more than it happened before.

What else though? Remember I have to actually want to write this list.

There’s another complaint about “remember all those things in that multi-agent system that is you, you’re just going to ignore what they all want and write down what the “sensible” one wants, with maybe a vague shoutout to one of the others that gets put at low priority”. What else?

Some kind of shame, a feeling that by writing that something needs doing on a list I need to go all over the feelings that I haven’t done it yet.

Anyway, I’m not sure this is going anywhere. Can’t I just ask the shadowy thing that’s in control what it wants?


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s