Toon Alfrink's sketchpad

by toonalfrink31st Oct 201971 comments
72 comments, sorted by Highlighting new comments since Today at 6:06 PM
New Comment
Some comments are truncated due to high volume. (⌘F to expand all)Change truncation settings

Case study: A simple algorithm for fixing motivation

So here I was, trying to read through an online course to learn about cloud computing, but I wasn't really absorbing any of it. No motivation.

Motives are a chain, ending in a terminal goal. Lack of motivation meant that my System 1 did not believe what I was doing would lead to achieving any terminal goal. The chain was broken.

So I traversed the chain to see which link was broken.

  • Why was I doing the online course? Because I want to become better at my job.
    • Do I still think doing the online course will make me better at my job? Yes I do.
    • Do I want to get better at my job? Nah, doesn't spark joy.
  • Why do I want to get better at my job? Because I want to get promoted.
    • Do I still think doing better will make me get promoted? Yes I do.
    • Do I want to get promoted? Nah, doesn't spark joy.
  • Why do I want to get promoted? Because (among other things) I want more influence on my environment, for example by having more money.
    • Do I still think promotion will give me more influence? Yes I do
    • Do I want more influence? Nah
  • Why do I want more influence (via money)? Because (among other things) I want to buy a house and do meetups, and live wi
... (read more)
3Matt Goldenberg8moAnother thing you can do when you get to that top level: "Is this the best way to get that? (living with close friends at the center of a vibrant community), if not, what is?"

Ibogaine seems to reset opiate withdrawal. There are many stories of people with 20 year old heroin addictions that are cured within one session.

If this is true, and there are no drawbacks, then we basically have access to wireheading. A happiness silver bullet. It would be the hack of the century. Distributing ibogaine + opiates would be the best known mental health intervention by orders of magnitude.

Of course, that's only if there are no unforeseen caveats. Still, why isn't everybody talking about this?

4romeostevensit9moAnti-tolerance drugs seem neglected, tractable, and scalable. We've done some shallow investigation at QRI and think it is pretty promising. Have been keeping it as a bullet point as we ask around in funding and academic circles. It's an area that could use a dedicated effort for sure.
4Swimmer9639moBased on a quick glance at the Wikipedia page, it looks like ibogaine may have a significant risk of toxicity (and also the experience of being on it does not sound necessarily fun? I would not choose to take it): [] Also, I think this would rely on opiates being a pleasure-causing experience for everyone or almost everyone, which doesn't seem obviously true to me. (Source: recently had major surgery, had experience of various opiates including given by IV, kind of hate all of them except for the part where they result in less physical pain.)
2romeostevensit9moWe can fund the search for analogs to ibogaine with no side effects. We can figure out if microdosing ibogaine works (some promising but weak evidence it does). Typical mind fallacy: people with good lives underestimate drastically how important low dose opiates can be for helping people with unlivable chronic pain.
2Swimmer9639moOh, yeah, for the subpopulation where opiates are therapeutic this seems really valuable. (Which, who knows, I could end up being in, if I'm unlucky and get chronic nerve pain from my amputation). But IMO that's a pretty different thing from "wireheading" or a "happiness silver bullet" and it really confuses the issue to call it that.
-1toonalfrink9moHave you tried opiates? You don't need to be in pain for it to make you feel great
4Swimmer9638moI had an above knee amputation due to cancer in March and have been on opiates – several different kinds, less over time, sometimes when I was in a lot of pain and sometimes more prophylactically when I'm not in pain but am preparing for something I expect to be painful. I mostly hate the experience of being on them, especially the "high" if I take it before I'm actually in pain from physiotherapy or whatever. (I do appreciate being in less pain. Pain is bad.) I...guess it's interesting and I could see a different person liking the experience? I get a lot of dissociative effects, especially with the IV opiates they gave me in hospital. (Feeling like I'm floating above my body, feeling like I don't have free will and am just watching my actions happen from a distance.) I don't particularly enjoy this. They also make me feel tired and out of it / cognitively impaired, and I am really, really averse to that. I ended up drinking so much coffee in the hospital trying to fight this off. My guess is that brains vary and some people would experience this as "feeling great". (I've noticed this with other things like stimulants; I really like how coffee makes me feel, for example, but I know a lot of people who experience it as anxiety/unpleasant jitteriness.)
2toonalfrink9moAs for being on ibogaine, a high dose isn't fun for sure, but microdoses are close to neutral and their therapeutic value makes them net positive

Today I had some insight in what social justice really seems to be trying to do. I'll use neurodiversity as an example because it's less likely to lead to bad-faith arguments.

Let's say you're in the (archetypical) position of a king. You're programming the rules that a group of people will live by, optimizing for the well-being of the group itself.

You're going to shape environments for people. For example you might be running a supermarket and deciding what music it's going to play. Let's imagine that you're trying to create the optimal environment for people.

The problem is, since there is more than one person that is affected by your decision, and these people are not exactly the same, you will not be able to make the decision that is optimal for each one of them. If only two of your customers have different favourite songs, you will not be able to play both of them. In some sense, making a decision over multiple people is inherently "aggressive".

But what you can do, is reduce the amount of damage. My understanding is that this is usually done by splitting up the people as finely as possible. You might split up your audience i... (read more)

4ChristianKl1yThe problem is that reducing the amount of damage is not the same thing as maximizing value. It's not utilitarian. If you are faced with making a decision that gives 99% of people +1 utility but 1% -10 utility, an approached that targets damage reduction means that you make a choice to leave 0.89 utility for the average person on the table. Giving someone consideration is not enough from a critical theory perspective. Consideration is not equity. Equity is about also giving them power to take part in the decision making.
3Viliam1yThis would be nice. But in practice I don't see splitting the audience along many dimensions; rather the differences are shoehorned into sex/gender, sexual orientation, and race (e.g. insisting that "Muslim" is a race). In a social justice debate, an asperger is more likely to be called an asshole than accepted as a disadvantaged minority. Also, the dimension of wealth vs poverty is often suspiciously missing. If you are a benevolent dictator, it would better to simply have two supermarkets -- one with music and one without -- and let everyone choose individually where they prefer to shop. Instead of dividing them into categories, assigning the categories to shops, then further splitting the categories into subcategories, etc. But this means treating people as individuals, not as categories. Specifically, trying to help people by helping categories is an XY problem [] (you end up taking resources from people at the bottom of the "advantaged" categories, and giving them to people at the top of the "disadvantaged" categories; for example Obama's daughters would probably qualify for a lot of support originally meant for poor people). Epistemically, social justice is a mixed bag, in my opinion. Some good insights, some oversimplifications. Paying attention to things one might regularly miss, but also evolving its own set of stereotypes and dogmas. It is useful as yet another map in your toolbox, and harmful when it's the only map you are allowed to use.
3G Gordon Worley III1yI find this interesting as this gives one of the better arguments I can recall for there being something positive at the heart of social justice such that it isn't just one side trying to grab power from another to push a different set of norms, since that's often what the dynamics of it look like to me in practice, whatever the intent of social justice advocates, and I find such battles not compelling (why grant one group power rather than another, all else equal, if they will push for the things they want to the exclusion of those who would then not be in power just the same as those in power now do to those seeking to gain power?).
2[comment deleted]1y

You may have heard of the poverty trap, where you have so little money that you're not able to spend any money on the things you need to make more. Being broke is an attractor state.

You may have heard of the loneliness trap. You haven't had much social interaction lately, which makes you feel bad and anxious. This anxiety makes it harder to engage in social interaction. Being lonely is an attractor state.

I think the latter is a close cousin of something that I'd like to call the irrelevance trap:

  • Lemma 1: having responsibilities is psychologically empowering. When others depend on your decisions, it is so much easier to make the right decision.
  • Lemma 2: being psychologically empowered makes it more likely for you to take on responsibility, and for others to give you responsibility, because you're more able to handle it.

I speculate that some forms of depression (the dopaminergic type) are best understood as irrelevance traps. I'm pretty sure that that was the case for me.

How do you escape such a trap? Well you escape a loneliness trap by going against your intuition and showing up at a party. You escape an irrelevance trap by going against your intuition and taking on more responsibility than you feel you can handle.

2Dagon1yI like this direction of thought. Note that for all of these traps, success is more often a matter of improvement rather than binary change or "escape from trap". And persistence plays a large role - very few improvements come from a single attempt.
2moses1ySounds very close to what Peterson says
1toonalfrink1yHe does influence my thinking

I did all the epistemic virtue. I rid myself of my ingroup bias. I ventured out on my own. I generated independent answers to everything. I went and understood the outgroup. I immersed myself in lots of cultures that win at something, and I've found useful extracts everywhere.

And now I'm alone. I don't fully relate to anyone in how I see the world, and it feels like the inferential distance between me and everyone else is ever increasing. I've lost motivation for deep friendships, it just doesn't seem compatible with learning new t... (read more)

4Dagon9moCan't speak for anyone else, and I don't know what my counterfactual selves feel like. [] - I don't know if you and I are similar in ways that matter on this topic. In fact, I don't know what mental features are important for how to optimize on this topic. Anyway, this is not advice, simply a framing that works for me. For me, I believe it's worth it. The tunnel widens a lot, and has LOTS of interesting features in it, but it does not end - there is a fairly fundamental truth underlying that loneliness, and I don't know of any acceptable ways for me to deny or forget that truth (to myself). I've become hyper-aware of the complexity and variance in humanity, and in myself moment-to-moment and year to year. This makes me quite able to have deep connections with many people, EVEN WHILE understanding that they model the universe differently than I on many dimensions. We can't have and don't need agreement on everything, or even on ontologically fundamental topics. We can agree that sitting around a campfire talking about our human experiences is desirable, and that's enough. With other groups, I can explore moral philosophy without a realism assumption, even if I don't particularly want to hang out with them on less intellectual topics.
2elityre8moI have often felt similarly.
2romeostevensit9moThe sense of ungroundedness sine waves over time afaict. The old strategies for connection had untenable foundations (e.g. tacit shared metaphysical assumptions), so you'll need to learn new ones. The Charisma Myth and NVC are good for bootstrapping some of the skills. Motivation in the new regime can't arise because you don't have even the proto-skills necessary to get a rewarding feedback loop going yet.

Here's a faulty psychological pattern that I recently resolved for myself. It's a big one.

I want to grow. So I seek out novelty. Try new things. For example I might buy high-lumen light bulbs to increase my mood. So I buy them, feel somewhat better, celebrate the win and move on.

Problem is, I've bought high-lumen bulbs three times in my life now already, yet I sit here without any. So this pattern might happen all over again: I feel like upgrading my life, get this nice idea of buying light bulbs, buy them, celebrate my win and move on.

So he... (read more)

2Raemon1yPresumably not the main point, what ends up happening to your luminators?
3toonalfrink1yMoved to a new country twice, they broke once. But the real cause is that I didnt regard these items as my standard inventory, which I would have done if I had more of a preservation mindset.

I have gripes with EA's that try to argue about which animals have consciousness. They assume way too readily that consciousness and valence can be inferred from behavior at all.

It seems quite obvious to me that these people equate their ability to empathize with an animal with the ability for the animal to be conscious, and it seems quite obvious to me that this is a case of mind projection fallacy. Empathy is just a simulation. You can't actually see another mind.

If you're going to make guesses about whether a species is conscious, you sho... (read more)

4Matthew Barnett1yNot to be pedantic, but what else could consciousness possibly be, except for a way of describing the behavior of some object at a high level of abstraction? If consciousness was not a behavior, but instead was some intrinsic property of a system, then you run into the exact same argument [] that David Chalmers uses to argue that philosophical zombies are conceivable. This argument was forcefully rebutted [] in the sequences. ETA: When I say behavior, I mean it in the physical sense. A human who is paralyzed but nonetheless conscious would not be behaviorally identical to a dead human. Superficially yes, but behavior means more than seeing what goes on outside. While you might say that I'm simply using a different definition of behavior than you were, I think it's still relevant, because any evolutionary reason for consciousness must necessarily show up in observational behavior, or else there is no benefit and we have a mystery.
1TAG1yIt could be something that is primarily apparent to the person that has it. That runs together two claims: that consciousness is not behaviour, and that it is independent of physics. You don't have to accept the second claim in order to accept the first. And it remains the case that Chalmers doesn't think zombies are really possible. "Primarily accessible to the person that has it" does not mean "no behavioural consequences".
2Matthew Barnett1yI'm not convinced that this definition is sufficiently clear, or that consciousness should be defined this way. Rather, it's a property of consciousness that people claim that it's "readily apparent", but I am not convinced that it has this readily apparent quality to it. In general, rather than taking the properties of consciousness at face value, I take Dennett's approach of evaluating people's claims about consciousness from a behavioral science perspective. From my perspective, once you've explained the structure, dynamics, and the behavior, you've explained everything. Are you sure? Chalmers argues in Chapter 3 of The Conscious Mind that zombies are logically possible. I am not really even sure what force the Zombie argument could hold if he thought it was not logically possible.
1TAG1yIt's not intended to be a complete definition of consciousness, just a nudge away from behaviourism. From my point of view, that's missing the central point quite badly. And elsewhere that they are metaphysically impossible.
5Matthew Barnett1yCould you let me know why? What about consciousness is missed by a purely behavioral description of an object (keeping in mind that what I mean by behavior is very broad, and includes things like the behavior of electrical signals)?
1TAG1yWhat is missed is the way it seems from the inside,as I pointed out originally. I don't have to put my head into an FMRI to know that I am conscious.
2Isnasene1yI think people who refer to animal behavior in making statements about consciousness are making a claim more along the lines of "given that a being has a brain with superficial similarities to ours and was evolved via a process similar to our own evolution, we can take it's behavior as higher level indicators of what its brain is doing and infer things about consciousness." Otherwise, these people would also grant consciousness to all sorts of things we make with superficially human behavior but obviously different mechanisms (ie non-playable characters in MMOs, chatbots). I read a lot more about consciousness back in the day and I'm not convinced that neural correlates are any better evidence for consciousness than behavior, given that the beings we're considering already have brains. I'm no expert but per wikipedia on neural correlations of consciousness, we don't have much in terms of neural correlates: Per Open Philanthropy's 2017 report [] on consciousness on cortex-requiring views (CRVs), we're not really sure how important having a cortex is for consciousness: And from the same report, there aren't really any clear biological factors that can be used to draw lines about consciousness: Moreover, people who have done way more thorough research into correlates of consciousness [] than me use both (ie anatomical features [] as an example of neural correlates, motivational trade-offs [] , as an example of behavior). Given that animals already have a bunch of similarities to humans, it strikes me as a mistake not to consider behavior at all.
2Dagon1yA reductio ad absurdum for this is the strong skeptical position: I have no particular reason to believe that anything is conscious. All configurations of quantum space are equally valuable, and any division into "entities" with different amounts of moral weight is ridiculous.
2Lanrian1yThe strong version of this can't be true. You claiming that you're conscious is part of your behaviour. Hopefully, it's approximately true that you would claim that you're conscious iff you believe that you're conscious. If behaviour doesn't at all correlate with consciousness, it follows that your belief in consciousness doesn't at all correlate with you being conscious. Which is a reductio, because the whole point with having beliefs is to correlate them with the truth.
1toonalfrink1yRight, right. So there is a correlation. I'll just say that there is no reason to believe that this correlation is very strong. I once won a mario kart tournament without feeling my hands.

Here's an idea: we hold the Ideological Turing Test (ITT) world championship. Candidates compete to pass an as broad range of views as possible.

Points awarded for passing a test are commensurate with the amount of people that subscribe to the view. You can subscribe to a bunch of them at once.

The awarding of failures and passes is done anonymously. Points can be awarded partially, according to what % of judges give a pass.

The winner is made president (or something)

2MakoYass3moWhy aren't presidential races already essentially ITT Tournaments? It would seem like that skill would make you really good at drawing support from lots of different demographics.
2ChristianKl3moThat basically reads like "just so you know, I'm kidding".
4toonalfrink3moIt's supposed to read like "this idea is highly unpolished"
-4ChristianKl3moOrganizing tournaments is something that requires little resources. Appointing presidents on the other hand requires resources that are fully out of reach. When talking about colonizing Mars speaking about using warp drives to get there doesn't read like "the idea is highly unpolished".
2toonalfrink3moLook, if you can't appreciate the idea because you don't like it's delivery, you're throwing away a lot of information
0ChristianKl3moGiven specific criticisms to proposals that people make isn't throwing away information.
4Matt Goldenberg3moIt seemed like Toon was trying to say "I have a cool idea for a tournament but haven't thought about a good prize"
2ChristianKl3moPrizes are not constraints for running tournaments and treating them like they are makes it harder to bring a tournament to reality.
2Matt Goldenberg3moI guess it depends on why you want to run the tournament.

As someone who never came across religion before adulthood, I've been trying to figure it out. Some of it's claims seem pretty damn nonsensical, and yet some of it's adherents seem pretty damn well-adjusted and happy. The latter means there's gotta be some value in there.

The most important takeaway so far is that religious claims make much more sense if you interpret them as phenomenological claims. Claims about the mind. When buddhists talk about the 6 worlds, they talk about 6 states of mood. When christians talk about a coven... (read more)

So here's two extremes. One is that human beings are a complete lookup table. The other one is that human beings are perfect agents with just one goal. Most likely both are somewhat true. We have subagents that are more like the latter, and subsystems more like the former.

But the emphasis on "we're just a bunch of hardcoded heuristics" is making us stop looking for agency where there is in fact agency. Take for example romantic feelings. People tend to regard them as completely unpredictable, but it is actually possible to predict to some extent whether y

... (read more)
5[anonymous]1yNeither one of those assumptions are true. There’s a lot we don’t know about neuroscience, but we do know that we’re not giant lookup tables, and we don’t have singular goals.
2Matt Goldenberg1y(Citation needed) I think both of those assumptions are unlikely, but am skeptical of your certainty.
2[anonymous]1yFor a citation, go open any neuroscience textbook. We are made of associative memories, but we are not one giant associative look-up table. These are vastly different architectures, and the proof is simply that neurons are locally acting. For humans to be a giant lookup table would require neural networking that we objectively do not have. For the claim about being singular goal systems, I could point to many things but maybe Maslow's hierarchy of needs (hierarchy of separate, factorable goals!) is sufficient:'s_hierarchy_of_needs ['s_hierarchy_of_needs]
2Matt Goldenberg1yIt's interesting because Maslow's Hierarchy actually seems to point to the exact opposite idea to me. It seems to point to the idea that everything we do, even eating food, is in service of eventual self-actualization. This is of course ignoring the fact that Maslow seems to basically be false experimentally.
2[anonymous]1yPrioritization of goals is not the same as goal unification. Citation?
2Matt Goldenberg1yIt can be if the basic structure is "I need to get my basic needs taken care of so that I can work on my ultimate goal". I think Kaj has a good link on experimental proof for Maslow's Hierarchy. I also think that it wouldn't be a stretch to call Self-determination theory a "single goal" framework, that goal being "self-determination", which is a single goal made up of 3 seperate subgoals, which crucially, must be obtained together to create meaning (if they could be obtained seperately to create meaning, and people were OK with that, than I don't think it would be fair to categorize it as a single goal theory.
2[anonymous]1yThat's a fully generic response though. Any combination of goals/drives could have a (possibly non-linear) mapping which turns them into a single unified goal in that sense, or vice versa. Let me put it more simply: can achieving "self-determination" alleviate your need to eat, sleep, and relieve yourself? If not, then there are some basic biological needs (maintenance of which is a goal) that have to be met separately from any "ultimate" goal of self-determination. That's the sense in which I considered it obvious we don't have singular goal systems.
2Matt Goldenberg1yYeah, I think that if the brain in fact is mapped that way it would be meaningful to say you have a single goal. Maybe, it depends on how the brain is mapped. I know of at least a few psychology theories which would say things like avoiding pain and getting food are in the service of higher psychological needs. If you came to believe for instance that eating wouldn't actually lead to those higher goals, you would stop. I think this is pretty unlikely. But again, I'm not sure.
2Kaj_Sotala1yThis BBC article [] discusses it a bit: Of course, this doesn't really contradict your point of there being separable, factorable goals. AFAIK, the current mainstream model of human motivation and basic needs is self-determination theory [], which explicitly holds that there exist three separate basic needs:
2[anonymous]1yThanks, I learned something. Although for the purposes of this discussion it seems that Maslow's specific factorization of goals is questionable, but not the general idea of a hierarchy of needs. Does that sound reasonable?
2Kaj_Sotala1yWell, it sounds to me like it's more of a heterarchy [] than a hierarchy, but yeah.
2Dagon1yThese aren't contradictory or extremes of a continuum, they're different levels of description of agency. A complete enough lookup table is indistinguishable from having goals. A deep enough neural net (say, a brain) is a pretty complete lookup table. The "one goal" idea is a slightly confused modeling level - "goal" isn't really unitary or defined well enough to say whether a set of desires is one conjunctive goal or many coordinated and weighted goals.

I've been trying to figure out what finance really is.

It's not resource allocation between different people, because the intention is that these resources are paid back at some point.

It's rather resource re-allocation between different moments in one person's life.

Finance takes money from a time-slice of you that has it, and gives it to a time-slice of you that can best spend it.

Optimal finance means optimal allocation of money across your life, regardless of when you earn it.

2Wei_Dai1yThat's part of it, but there's also coordination between people (e.g., investors coming together to finance a capital-intensive business that no single person can afford to fund) and managing risk and incentives (e.g., a sole proprietorship has better incentives but worse risk characteristics compared to a company with many shareholders, a company issuing both stocks and bonds so investors can make their own risk/reward tradeoff, etc.). I think maybe something like "finance is about how a group of people can cooperate to pursue EU maximization over time, given some initial endowment of assets and liabilities" would capture most of it?
1toonalfrink1yWell yes, but no, the point is that these other people are merely means, but optimally distributing your assets over time is a means that screens off the other people, in a sense. In the end, assuming that people are really just optimizing for their own value, they might trade for that but in the end their goal is their own allocation.
2Wei_Dai1yOk, I think that makes sense.
2Dagon1yWhen exploring things like this, taboo the phrase "really is". No abstraction really is anything. You might ask "how can I use this model", or "what predictions can I make based on this" or "what was that person trying to convey when they said that". Less philosophically, you're exploring an uncommon use of the word "finance". Much more commonly, it's about choice of mechanism for the transfer, not about the transfer itself. What you describe is usually called "saving" as opposed to "consumption". It's not finance until you're talking about the specific characteristics of the medium of transfer. Optimal savings means good choices of when to consume, regardless of when you earn. Optimal finance is good placement of that savings (or borrowing, in the case of reverse-transfer) in order to maximize the future consumption.
1toonalfrink1ySure. In this case "what it really is" means "what does it optimize for, why did people invent it"
2Hazard1yI think putanumonit wrote a post []that made a similar point. Might be useful in this investigation.

Personalized mythic-mode rendition of Goodhart's law:

"Everyone wants to be a powerful uncompromising force for good, but spill a little good and you become a powerful uncompromising force for evil"

Question for the Kegan levels folks: I've noticed that I tend to regress to level 3 if I enter new environments that I don't fully understand yet, and that this tends to cause mental issues because I don't always have the affirmative social environment that level 3 needs. Do you relate? How do you deal with this?

4Matt Goldenberg8moBy dealing with trauma and taking shame and guilt as object. By incorporating the need for belonging into my systemic understanding. I had this for a long time but it's pretty rare now.
[+][comment deleted]7mo 2