Posts
Comments
(black) hat tip to johnswentworth for the notion that the choice of boundary for the agent is arbitrary in the sense that you can think of a thermostat optimizing the environment or think of the environment as optimizing the thermostat. Collapsing sensor/control duality for at least some types of feedback circuits.
I refer to a version of this as conspiracy theory thinking. At any given time you have several problems you are working on, across various domains of your personal and professional life. The fake framework is that they're all secretly generated by the same problem, but there's a conspiracy to make them seem like different problems. Your job is to unravel the conspiracy.
I think it's that any basis set I define in a super high dimensional space could be said to be value laden, though it might be tacit and I have little idea what it is. If I care about 'causal structure' or something that's still relative to the sorts of affordances that are relevant to me in the space?
First linky no worky.
Signed up.
Your self model only contains about seven moving parts.
Your self model's self model only contains one or two moving parts.
Your self model's self model's self model contains zero moving parts.
Insert UDT joke here.
Thanks for writing this up and I also wanted to say I just really like the phrase 'kissing scars.'
Several years on, I've strongly updated in the direction that this is neglected and underrated both for health and happiness. I now do yoga a bunch in addition to the above.
Meta: https://www.lesswrong.com/posts/37sHjeisS9uJufi4u/scholarship-how-to-do-it-efficiently
Object: https://www.lesswrong.com/posts/bZ2w99pEAeAbKnKqo/optimal-exercise
The need for ongoing maintenance is a feature not a bug given that we don't understand the actual effects in high resolution.
Between this and the viability of other geoengineering projects I'm expecting that the biggest blocker will actually be how to make sure we don't overshoot in the other direction, or just introduce chaotic variance.
Next steps are connecting an interested funder, the most competent researcher in the area, and someone who has some relevant engineering knowledge to figure out which experiments would maximize further information to tackle the biggest uncertainties in the model. Mainstream philanthropy will only take interest once someone does the weird thinking and logistics for them.
I would guess ~zero chance of any public funding for this. Professional hand wringers aren't incentivized to solve problems. If before any charities had already scaled it you had approached people and told them you thought distributing insecticide sprayed bed netting might be the most important thing in global welfare you would have gotten a mix of polite interest and scoffing that never would have gone anywhere.
Edit: looks like I'm wrong. The project is housed at a public university which presumably means it has received some grants. I'd still expect a motivated team to outperform the entire org. Looks like they produce <1 paper per year on average. All the collaborators probably have other things that are their main projects.
It's worth noting that projected costs are as high as 1 billion per year per W/m^2 of reduced forcing. Still trivial in the grand scheme of things, but quite a bit different than expected total cost of 10B.
Things like Focusing and goal factoring are a good first pass.
It seems to me that that piece has to do a lot of scaffolding because it doesn't use the compression of 'degrees of freedom.' Eg your explanans has to have fewer degrees of freedom than the explanandum.
The 'art' in picking the correct theorem in B seems related to structural realism. ie figuring out where we are importing structure from and how as we port across representations.
learn more from. Edited.
Current events are hard to extract signal from.
I agree but found separating everything out a lot of work.
- 1. Which questions are important?
- a. How should we practice cause prioritization in effective altruism?
- Encourage people to follow different prioritization heuristics and see what bubbles up. Funding people who are doing things differently from how you would do them is incredibly hard but necessary. EA should learn more from Jessica Graham.
- b. How should we think about long shots at very large effects? (Pascal's Mugging)
- Seems to vary based on risk appetite and optionality. ie young people can do moonshots and recover in time to do other lower variance things.
- c. How much should we be focusing on the global level, vs. our own happiness and ability to lead a normal life?
- false dilemma. Inquiring into values shows that focusing on the global level contributes to my happiness. 'Lead a normal life' seems to be about priors on 'normal behaviors' leading to well being. But this prior seems bad, average outcomes aren't very happy.
- d. How do we identify gaps in our knowledge that might be wrong and need further evaluation?
- More focus on critiques that induce physical discomfort and avoidance.
- e. How do we identify unexamined areas of our lives or decisions we make automatically? Should we examine those areas and make those decisions less automatically?
- per Buddhism, probably. For intellectual types the how is often somatic skill training.
- 2. How do we determine whether we are operating in the right paradigm?
- a. What are paradigms? Are they useful to think about?
- a paradigm is a collection of heuristics that play well together. ie they chain into each other easily by taking each others outputs as inputs.
- b. If we were using the wrong paradigm, how would we know? How could we change it?
- Observing the outcomes of the people following different stances. I think Opening the Heart of Compassion is an excellent resource here, as is The Five Personality Patterns. Despite woo.
- c. How do we learn new paradigms well enough to judge them at all?
- Ask ourselves the question 'do I want that person's life?' When evaluating strategies.
- 3. How do we determine what the possible hypotheses are?
- a. Are we unreasonably bad at generating new hypotheses once we have one, due to confirmation bias? How do we solve this?
- b. Are there surprising techniques that can help us with this problem?
- Creativity techniques around prioritizing quantity over quality works reliably.
- 4. Which of the possible hypotheses is true?
- a. How do we make accurate predictions?
- By creating the conditions for the outcomes we want. Too many degrees of freedom and hidden variables when trying to predict useful things totally outside our control. Collecting better outside view search heuristics for things we're forced to make predictions for that we can't control.
- b. How do we calibrate our probabilities?
- Practice feeling the somatic difference between 60 and 70% confidence, increase granularity with time.
- 5. How do we balance our explicit reasoning vs. that of other people and society?
- a. Inside vs. outside view?
- Always generate an inside view the best way you know how first. Then when you run an outside view and encounter differences, inquire into the generators of those differences. This is free calibration data any time you're about to search for something.
- b. How do we identify experts? How much should we trust them?
- Judgmental bootstrapping and checking how granular the feedback the expert has received in forming their model. Granularity of feedback should be >= granularity of decision model.
- c. Does cultural evolution produce accurate beliefs? How willing should we be to break tradition?
- Try harder to learn from tradition than you have been on the margin. Current = noisy.
- d. How much should the replication crisis affect our trust in science?
- It should increase our need for consilience in order to be confident about anything. If a conclusion isn't reachable from radically different methods/domains it's fairly suspect.
- e. How well does good judgment travel across domains?
- Granularity of feedback loops applies again. When we see impressive transfer I think it's from a domain with good feedback loops to a domain with poor feedback loops where the impressive person helped clean up those poor feedback loops by applying the methods from the high feedback loop domain.
- 6. How do we go from accurate beliefs to accurate aliefs and effective action?
- a. Akrasia and procrastination
- b. Do different parts of the brain have different agendas? How can they all get on the same page?
- Integrate conflicting parts via some psychotherapy modality like Focusing or Core Transformation or IFS.
- 7. How do we create an internal environment conducive to getting these questions right?
- a. Do strong emotions help or hinder rationality?
- emotional 'strength' seems like the wrong frame.
- b. Do meditation and related practices help or hinder rationality?
- Help on cognitive reflection tests and the general skill of noticing which cognitive heuristics are currently being run. CRTs were among the less correlated with g according to The Rationality Quotient.
- c. Do psychedelic drugs help or hinder rationality?
- Ultra high openess without skepticism/disagreeableness/epistemic hygiene seems to result in loopy beliefs. They should be leveled up in tandem.
- 8. How do we create a community conducive to getting these questions right?
- a. Is having "a rationalist community" useful?
- b. How do strong communities arise and maintain themselves?
- c. Should a community be organically grown or carefully structured?
- d. How do we balance conflicting desires for an accepting where everyone can bring their friends and have fun, vs. high-standards devotion to a serious mission?
- e. How do we prevent a rationalist community from becoming insular / echo chambery / cultish?
- f. ...without also admitting every homeopath who wants to convince us that "homeopathy is rational"?
- g. How do we balance the need for a strong community hub with the need for strong communities on the rim?
- h. Can these problems be solved by having many overlapping communities with slightly different standards?
- I think communities are typically about avoiding responsibility for making personal progress. People who choose to take a more central role in a community typically have emotional problems they are trying to work out via the dynamics in the community. The whole is typically much less than the sum of its parts.
- 9. How does this community maintain its existence in the face of outside pressure?
- The way in which outside pressure is experienced is worth investigating for what internal process it is resonating with.
Lack of a 'recognition' state increases ambiguity in goal directed tasks i.e. no clear picture of what an organized cupboard looks like generates friction in beginning the task due to lack of a feedback mechanism. In contrast, a clear internal picture generates lots of intermediate states to compare to.
People able to work on X have reasonably good proxy measures so that they can feel their way through the ambiguous parts of problems.
> When humans made advances in coordination ability in the past, how was that accomplished? What are the best places to apply leverage today?
I am confused by the general lack of interest I've encountered in how joint stock corporations came to be and underwent selection to get us to where we are now. It may be I'm not looking in the right places. I know the founders of Mckinsey are quite interested in this.
thanks!
Right, I'm saying I started off con, you started off pro, and both of us nudged towards the other along the dimensions mentioned. I probably updated more than you.
This was more than 2 years ago so unfortunately I don't recall.
Today I wondered how much of forecasting plus data science is just people realizing that insurance isn't the only use of actuarial methods.
John Maxwell and I went back and forth on this question a bunch. I was initially on the con side and updated in the direction that increasing lurker ratios and common knowledge generation are big enough considerations that forum participation is probably pretty good/helpful. He was initially on the pro side and I think updated in the direction that without an anchor of scholarship (textbooks, research review, and structured notes) and longer/more in depth conversations online activity can acquire a veneer of productivity that is harmful to both real productivity and mental health.
In metaphysical conflicts people don't win by coming up with the best evidence, they win by controlling what gets counted as evidence. By default, memeplexes gain stability by creating an environment in which evidence against them can't be taken seriously. Arguments that EA has failed to actually measure the things it claims are worth measuring should be taken very seriously on their face, since that is core to the claims of moral obligation (which is itself a bad frame, but less serious.)
You already had the theory of mind link
I take anti inflammatories though I haven't blinded subjective effects. Berberine (similar in action and effect size to metformin), ginger, curcumin phytozome. Rest is the same. D3+k2. Theanine + ashwaghanda if I'm having a stimulant. Melatonin if sleep is disrupted. If one gets a subjective effect from a supplement one should investigate why this might be the case and try to solve the root cause. ie for me I had above normal inflammation from foods I didn't know I was allergic to (mild).
Since most papers don't include much metadata this will be really hard to figure out (also, which citations count as central to the insight?). I agree knowing the answer to this would be very interesting. My impression has been that Kuhnian style shifts do generally involve someone going back to the assumptions of the current paradigm and realizing that a different direction is now plausible given what has been discovered in the interim. E.g. the modern era is built on set theory and point estimates. In order to make progress rebasing on distributions natively might have to happen.
Allowing donors to coordinate better in the face of uncertainty about information sources seems really important even if it doesn't directly wind up applying to ML.
Cruxing on implicit predictions as revealed by life strategies is a good trust building exercise. Including examining the discomfort that comes up.
Explanations can be about a few different things. I've been having some luck splitting them into variant (who, when, where) and invariant (how, why, what) parts. Aristotle and Sun Tzu had their own type systems for causality, and I'm curious why I haven't been able to find much in either philosophy or computer science about it. One guess is that I just haven't found the right keywords yet, but I have gone hunting around citation chains from knowledge representation and Pearl's stuff, as well as the stuff from here and here. The modal logic stuff seems promising but most of what's been built on it seems like epicycles.
In the 'keep the organization from being overrun' sense, see also sealioning. The search space of worthwhile things is very large and idiosyncratically explored by well meaning, intelligent people. Aggressive value laden 'logical arguments' often point to a tacit value to have everyone converge on the same set of metaheuristics. This is because the person doing this has a strong need for internal consistency that they are externalizing onto their social space. And there's nothing wrong with wanting internal consistency. But if pressed hard, it is anti-truth seeking as an aggregate strategy because you lose out on the consilience of having different people pursuing different search methods. Epistemology is a team sport. The objection would be 'but if we don't then argue about what we've discovered what's the point?' The point is that adversarial processes as a part of the truth seeking process needs to be consensual. This applies doubly when you aren't in a 101 space and people might be sick of a dynamic where simple seeming questions with complicated answers make newer members feel entitled to the effort needed to explain said complicated answers. This is one of the reasons well written blog posts that can be referenced by name can be so helpful for community discourse.
I like this post by the way and my comment wasn't an objection to it.
In general: making somatic reactions part of the mutual understanding of what is happening. The more you are aware of when you're abstracting from a somatic reaction the more you'll recognize when others are doing it too. E-Prime is aimed at this, as the step of symbolizing one thing as another thing (X *is* Y) is a moment in which you can catch this.
I feel like the elephant in the room is that convincing logical arguments are often only weak to moderate evidence for something.
What does it mean to pay utility?
The bar that is set for appeals to consequences imply the sort of equilibrium world you'll end up in. Erring on the side of higher is better, because it is hard to go the other way because epistemic standards tend to slide in the face of local incentives.
I also want to note an argumentative tactic that occurs on the tacit level whereby people will push you into a state where you need to expend more energy on average per truth bit than they do, so they eventually win by attrition. Related to evaporative cooling. The subjective experience of this feels like talking to the cops. You sense that no big wins are available (because they have their bottom line) but big losses are, so you stop talking. If you've encountered this dynamic, you recognize things like this
> "You still haven't refuted my argument. If you don't do so, I win by default."
as part of the supporting framework for the dynamic and it will make you very angry...which others will then use as part of the dynamic which makes you angry which......
Slipping out of production possibility frontier zero-sum relations involves slipping sideways along some other dimension. Saddle points in gradient descent. This is yet another reason why 'If a problem seems hard, the representation is probably wrong' often holds.
Higher in more expensive markets, but yes.
Workflowy. Dynalist and others have more features, but I don't want more features. More features means more decisions. I organize by month and do a review at the end of months, plus a year end review when collapsing into my archive tab. Tags for things like book notes, quotes, routines etc.
Since The Rationality Quotient mostly showed that Rationality isn't much of a thing on top of g, that means that despite not caring about the quality of arguments too much, other people aren't suffering worse life outcomes. One can take it as an opportunity to be curious about why that might be. What might others who seem to be less explicitly/verbally committed to truth be getting right in other ways? I've found that spiritual communities are good for this, and more open to reflection than most, once the right semantic flags are understood and translated.
I call it decision leverage, for handiness, and looking for it changes casual conversation as well as analysis.
Human civilizations have survivorship bias such that you're seeing the iterated output of those that found parameter insensitive parts of critical distributions. Exogenous shocks are common, and civs that weren't metastable disappeared. This happened on both the cultural and genetic level (you are already a civilization of trillions of cells). As the pace of change accelerates things might get more dangerous as people push farther/harder within the span of a single lifetime before natural self corrections kick in (four horsemen). The current trend could be seen as the human race essentially tinkering with cancer (runaway, uncontrollable growth) to see if it will grant us immortality before it kills us.
I liked the vulnerability of admitting wanting attention and the exploration of multiple consequences of that, the way coping strategies evolve and change as a result of pressures.
I really appreciated this post.
What Liron said. A small number of days accounts for most stock gains. Sharpe concluded that you'd need to make calls right about 74% of the time to win by timing.
There are ways to systematically make money. Diversification, not timing, low time preference, leverage, not being loss averse and other factors allow you to harvest money off of the poorly diversified, the market timers, the high time preference, the leverage averse, and the loss averse over time.
The popular saying is that bond yield inversions have predicted 6 of the last 3 recessions. In general, just run your same searches and append the word 'myth', 'misconception', 'doesn't' etc. If you use positive search methods the internet will tell you whatever you want.
More generally, market timers lose hard. Set your allocations such that you're comfortable with max drawdowns and plug money into it reliably. Recessions last on average 9 months and have an average draw down of 20%, not much to worry about in the long run.
I think people have alarm fatigue due to insufficient resolution of their priority function to make decisions without tons of effort. There are a large number of things in the world that are deeply important, but of upstream importance is knowing that there's an order that will have to de facto be followed whether you are aware of the things that determine said order or not. And that you won't personally have time to get to almost any of the urgent things.
I'm pointing this out because lots of things shouting for attention with highest priority signals is a recipe for burnout, while at the same time we do in fact need people still capable of sitting bolt upright occasionally.
So is the bet for the construction of a financial instrument that moves like a leveraged case-shiller housing index for a selected city with tracking error + fees smaller by some factor than the average frictional costs in the buying market?
edit: oh in this model we're mostly concerned with the rent hedging value rather than the speculative value (that being seen as mostly a thing that subsidizes the rent hedge here) so we'd really want something that pays you when the rent index deviates substantially from its trend. If we were happy with just hedging the long term trend I think that could be done cheaply with fixed income levered appropriately, but I'd have to show that it won't suddenly change correlations under various conditions.
edit 2: I think there's some aether variable-ing going on here. Rental market are much more liquid and display lower volatility than housing markets. So in order to purchase a hedge against rental price (cash flow) appreciation we expose ourselves to much more volatile net worth swings? That only makes sense for people with poor savings (almost all 'normal' people. But normal people aren't trying to buy in the most expensive markets anyway).