Posts

What AI companies would be most likely to have a positive long-term impact on the world as a result of investing in them? 2020-09-21T23:41:24.281Z · score: 6 (3 votes)
If there were an interactive software teaching Yudkowskian rationality, what concepts would you want to see it teach? 2020-09-02T05:37:08.758Z · score: 22 (14 votes)
Inoculating against Psychedelic Woo 2020-08-21T05:54:57.784Z · score: 8 (4 votes)
MikkW's Shortform 2020-08-10T20:39:29.510Z · score: 2 (1 votes)
Calibrate words, not just probabilities 2020-07-18T05:56:11.120Z · score: 11 (3 votes)

Comments

Comment by mikkel-wilson on MikkW's Shortform · 2020-09-29T17:35:30.483Z · score: 1 (1 votes) · LW · GW

Wow, that video makes me really hate Peter Thiel (I don't necessarily disagree with any of the points he makes, but that communication style is really uncool)

Comment by mikkel-wilson on Viliam's Shortform · 2020-09-28T15:55:01.396Z · score: 1 (1 votes) · LW · GW

This seems likely to me, although I'm not sure "superstimulus" is the right word for this observation.

It certainly does make sense that people who are inclined to notice the general level of incompetence in our society, will be less inclined to trust it and rely on it for the future

Comment by mikkel-wilson on ryan_b's Shortform · 2020-09-28T15:25:28.612Z · score: 1 (1 votes) · LW · GW

But the bottom line is that the value of weapons is destruction

The bottom line is protection, expansion, and/or survival; destruction is only an intermediate goal

Comment by mikkel-wilson on Rudi C's Shortform · 2020-09-28T05:10:01.775Z · score: 1 (1 votes) · LW · GW

I'm thinking "project [/product] announcement". I encourage you to add a tag you think works, if anyone comes up with a better name, we can always change the name later

Comment by mikkel-wilson on Honoring Petrov Day on LessWrong, in 2020 · 2020-09-26T08:15:33.552Z · score: 3 (2 votes) · LW · GW

During Thursday 26th September (midnight to midnight Pacific Time), we will practice the skill of sitting together and not pressing harmful buttons.

Happy Thursday

Comment by mikkel-wilson on MikkW's Shortform · 2020-09-26T08:10:43.500Z · score: 9 (2 votes) · LW · GW

I'm quite baffled by the lack of response to my recent question asking about which AI-researching companies are good to invest in (as in, would have good impact, not necessarily most profitable)- It indicates either A) most LW'ers aren't investing in stocks (which is a stupid thing not to be doing), or B) are investing in stocks, but aren't trying to think carefully about what impact their actions have on the world, and their own future happiness (which indicates a massive failure of rationality)

Even putting this aside, the fact that nobody jumped at the chance to potentially shift a non-trivial amount of funding away from bad organizations and towards good organizations (which I'm investing primarily as a personal financial strategy), seems very worrying to me. While it is (as ChristianKI pointed out) debatable that the amount of funding I can provide as a single person will make a big difference to a big company, it's bad decision theory to model my actions as only being correlated with myself; and besides, if the funding was redirected, it probably would have gone somewhere without the enormous supply of funds Alphabet has, and very well could have made an important difference, pushing the margins away from failure and towards success.

There's a good chance I may change my mind in the future about this, but currently my response to this information is a substantial shift away from the LW crowd actually being any good at usefully using rationality instrumentally

Comment by mikkel-wilson on What AI companies would be most likely to have a positive long-term impact on the world as a result of investing in them? · 2020-09-22T15:55:03.980Z · score: 2 (2 votes) · LW · GW

Thanks for the answer, Christian

If this analysis is correct, that would update me in favour of wanting to invest in smaller, currently under-funded startups that are alignment-aware, over for established players like Alphabet or Microsoft.

Although I'm not completely convinced that this is completely the case- based on the difference in performance between GPT-2 and GPT-3, it seems to me that we are likely in a hardware overhang right now, which means that the question of who is in a dominant position to decide the future of the world via AGI, is largely determined by who is able to first rally enough hardware to scale current techniques to AGI scale (although this isn't the sole consideration, only a dominant one). That means that even a relatively minor change in funding could provide a company just enough access to hardware to let them develop AGI slightly before their competition does, which could have huge long-term effects depending on what (if anything) that AGI locks in.

Comment by mikkel-wilson on I'm Voting For Ranked Choice, But I Don't Like It · 2020-09-21T14:24:37.317Z · score: 2 (2 votes) · LW · GW

It's not clear to me that this lens is more valid than the opposite. Sure, passing up reform now could mean no reform for a while- but one could argue that in the absence of any reform, frustration and pressure is maintained, meaning the conditions for reform continue to exist, whereas doing inferior reform makes a large portion of people say either "Yes, we got reform, so now we should be happy", or "Hey, we already tried reform, and it ended terribly. Why are they demanding we try it again?"

Lock-in effects are real, and I am concerned that settling for IRV will lock a barely-better system in place, when the conditions are currently ripe for a truly decent system to be implemented

Comment by mikkel-wilson on Progress: Fluke or trend? · 2020-09-14T03:42:36.047Z · score: 4 (3 votes) · LW · GW

Things that have underperformed for decades almost never take off.

That is pretty much the exact opposite of how technology actually works.

"We've been trying human flight since Da Vinci, but after decades of trying, we can conclude humans will never fly"

"We've been trying to cure smallpox for decades, but it's still endemic, so we will never eradicate smallpox"

"We've been trying to turn lead into gold for centuries, so it must not be possible" (granted, this one is only physically possible now, I don't think it's economically viable, but my point stands)

Comment by mikkel-wilson on Progress: Fluke or trend? · 2020-09-14T03:31:19.997Z · score: 3 (2 votes) · LW · GW

If you believe that your side of the bet is highly correlated with civilizational collapse, an alternative would be for me to pay you $X upon agreeing to the bet, and then for you to pay $3X (= X + 2X) conditional on the bet resolving in my favor. Realistically, the amounts would be adjusted for the risk of not collecting, plus economic growth and inflation, so not exactly a 3x ratio.

As for stakes, what stakes would you consider to be "not chump change"?

Comment by mikkel-wilson on Progress: Fluke or trend? · 2020-09-13T23:30:21.906Z · score: 5 (3 votes) · LW · GW

Everything you're saying fits the common narrative; I just think there's a roughly 80 percent chance that it's wrong.

I'm happy to bet on this. I propose the following operationalization:

Solar energy will be said to be successful if there is at least one year prior to (or including) 2040 where the total solar energy generated (excluding indirect pathways [which includes fossil fuels and wind], but including all forms of directly turning sunlight, or the sun itself, into human-usable energy, including by photosynthesis-based agriculture) is equal to the entirety of energy consumption in 2017 (113,000 TWh according to wikipedia).

If solar energy is "successful" according to this criterion, then you would pay me $30, if not, then I would pay you $15 (this ratio should be more than fair if you believe there is an 80% chance my analysis is wrong - especially since there are scenarios where my analysis is still valid, but the bet resolves in your favour)

Comment by mikkel-wilson on Progress: Fluke or trend? · 2020-09-13T23:10:13.301Z · score: 1 (1 votes) · LW · GW

The current profile of where our energy comes from has nothing to do with the long term prospects. There's a lot of progress to be made with solar, especially regarding the price of modules (but also in light-to-electricity efficiency), while we're already near the peak of what we are able to accomplish with fossil fuels.

Edit: I'd also be remiss if I didn't emphasize what the 2.45 EROI (Energy return on investment) means - assuming that it is accurate and representative of the actual situation (which the nature of the source you provide doesn't inspire confidence that it is correct). That means for every joule of energy used to manufacture and install a solar panel, 2.45 joules are returned over the lifetime of the installation (nb: it's not clear to me that the source measured lifetime energy output). That means that in terms of energy, the panel represents a net gain, and even using nothing but solar panels as a source of energy, and assuming the figure you cite is correct, a civilization can easily bootstrap itself from a small civilization to a great, powerful civilization- and certainly being forced to fall back on a 2.45 EROI is far from causing a collapse of civilization as we know it (as some of your comments have implied). For things to get dicey, EROI would have to fall below 1 (arguably a value very close to 1 would cause a lot of turmoil, which might lead to destruction of value-generation, leading to a long-term decrease in welfare, but 2.4 is more than enough to recover from short-term setbacks)

The only reason why a 2.45 EROI would seem disappointing is from an economical perspective- oil rigs provide an EROI around ~10, so (ignoring externalities and potential benefits that may not be easily measured in terms of energy) it would be a stupid idea to invest in an energy source that offers 2.45 EROI when you have a 10x return available instead. But of course, externalities are a thing, and the externalities caused by fossil fuels are worse per joule than the externalities caused by solar; and the EROI of solar is expected to only go up, while the EROI of fossil fuels will be going down as they become more scarce- so at some point, there will come a point where the EROI of solar will be a good deal (and remember, 2.45 is still a good deal if the alternative is nothing- good enough to continue to improve the quality of our lives, build colonies on other celestial bodies, radically improve our transportation networks, and solve world hunger)

Comment by mikkel-wilson on Sunday September 13, 12:00PM (PT) — talks by John Wentworth, Liron and more · 2020-09-13T16:08:24.813Z · score: 1 (1 votes) · LW · GW

Will there be a hangout afterwards? If so, on which platform? I don't plan on listening to the talks today, since I need to do things (although today's talks sounds fun), but if there's a hangout, I'd love to be at that.

Comment by mikkel-wilson on Comparative advantage and when to blow up your island · 2020-09-13T15:28:29.680Z · score: 1 (1 votes) · LW · GW

In section IV Zopa, this text is repeated twice:

Of course, I’d prefer to pay you fewer bananas! So I’d prefer a rate to the left end of this range. Conversely, it takes you twice as long to make banana as a coconut. You’d be thrilled if I paid you 4 bananas per coconut, but you’d never accept less than 1/2 a banana for one coconut.

Comment by mikkel-wilson on Progress: Fluke or trend? · 2020-09-13T15:07:29.527Z · score: 3 (2 votes) · LW · GW

The "materialist" position is very short-sighted. There's more than enough energy in both nuclear and solar. We have the technical foundations to run the world on nuclear just as effectively as fossil fuels, and nuclear sources are more abundant (in total usable energy) than fossil fuels, so there's really no open question about nuclear. It's purely a matter of supply and demand now (i.e., as fossil fuels become more expensive / get a worse reputation, we will naturally transition to nuclear)

As for solar, again, more energy reaches the Earth in the form of sunlight each year than the entirety of humanity uses each year (and that's not even counting the sunlight that doesn't reach Earth. Only a tiny, tiny fraction of the sun's light reaches the Earth, so if we are able to utilize the entirety of the sun's energy, we have billions of times as much energy as we have now). We're right on the knife's edge of solar technology being viable, so it generally only gets deployed where it makes the most sense, but as solar technology gets cheaper, we will see solar be deployed in more creative and flexible ways, and will make up a larger and larger portion of our energy use

Comment by mikkel-wilson on MikkW's Shortform · 2020-09-12T01:58:10.078Z · score: 1 (1 votes) · LW · GW

The intention is that there would be not two separate companies, but one company which is split between being owned fully by the entrepreneur, and being managed by the entrepreneur- so the entrepreneur would still be motivated to make the company do as well as possible, thereby generating revenue for the public at large

Comment by mikkel-wilson on MikkW's Shortform · 2020-09-12T01:18:27.021Z · score: 1 (1 votes) · LW · GW

Huh, I'm seeing a 404 when I click the link

Comment by mikkel-wilson on MikkW's Shortform · 2020-09-09T19:33:46.755Z · score: 5 (3 votes) · LW · GW

Viliam's shortform posts have got me thinking about income taxes versus wealth taxes, and more generally the question of how taxes should be collected. In general I prefer wealth taxes over income taxes, although I suspect there may very well be better forms of taxes than either of those two - But considering wealth taxes specifically, I think the main problem with wealth taxes is that over the long term they take away control of resources from people who have proven in the past that they know how to use resources effectively, and while this can allow for short-term and medium-term useful allocations of resources, it prevents very long horizon investing – as exemplified by Elon Musk's projects including SpaceX, Tesla, Neuralink and The Boring Company – projects that are good investments primarily because Musk understands that in the very long term these projects will pay off – both in personal financial returns and in general global welfare. While Tesla is very close to becoming profitable (they could turn a profit this year if they wanted to), and SpaceX isn't too far off either, he founded companies without any eye for medium term profits - he founded them understanding the very long game, which is profitable in the absence of year-over-year wealth taxes, but could potentially be unprofitable if year-over-year wealth taxes were introduced

The proposal that came across my mind in regards to alleviating the negative impact wealth taxes would have this way, is to allow entrepreneurs to continue to have control of the money they pay in wealth taxes, but that money is held in trust for the greater public, not for the personal use of the entrepreneur.

To clarify my point, I think it's worth noting that there are two similar concepts that get conflated into the single word "ownership": the 1st meaning of "own" (personal ownership) is that a person has full rights to decide how resources are used, and can use or waste those resources for their own personal pleasure however they wish; the 2nd meaning of "own" (entrusted to) is that a person has the right to decide how resources are used and managed, but ultimately they make decisions regarding those resources for the good of a greater public, or another trustor (entrusting entity), not for themselves.

When resources are owned by (i.e., entrusted to) somebody, they have the right to allocate those resources however they think is best, and aside from the most egregious examples of the resources being used for the personal gain or pleasure of the trustee, nobody can or should question the judgement of the trustee.

Back to wealth taxes: in my proposal, while an entrepreneur would still be expected to "pay" a certain percentage of their wealth each year to the greater public, instead of the money going directly to the government, the resources will instead continue to be "owned" by the entrepreneur, but instead of being personally owned for the entrepreneur's gain and pleasure, it would be entrusted to the entrepreneur in the name of the public, and the entrepreneur will be allowed to continue to use the resources to support any enterprises they expect to be a worthwhile investment, but when the enterprise finally turns a profit, the percentage of revenues that correspond to the part that is entrusted in the name of the public, will then be collected as taxes.

The main benefit of this proposal (assuming wealth taxes are already implemented) is that, while it cannot make profitable any venture that would be rendered unprofitable by a wealth tax, it can maintain the feasibility of ventures that are profitable in the long run, but which are made unfeasible in the short and medium terms by a wealth tax, due to the cost of taxes being more than medium term gains.

Comment by mikkel-wilson on Sunday September 6, 12pm (PT) — Casual hanging out with the LessWrong community · 2020-09-06T22:29:21.639Z · score: 1 (1 votes) · LW · GW

After having experienced both Gather.town and Topia, I find Topia to feel better than Gather. Even though it seems like a small thing, the grid based movement in Gather just feels too chunky, and the mobility in Topia feels a lot more natural than in Gather

Comment by mikkel-wilson on If there were an interactive software teaching Yudkowskian rationality, what concepts would you want to see it teach? · 2020-09-04T16:50:05.207Z · score: 1 (1 votes) · LW · GW

I agree that Syntorial has better interactivity than many of the "explorables" that have become popular lately, and I agree that high interactivity is vital for maximizing learning.

As far as the actual implementation of Syntorial, beyond the fact that it succeeds at having high interactivity, I find that the user experience lacks flow, and I find it fairly unengaging - in particular the videos slow down the pace, and I would generally want to skip them, but often don't because I worry about missing important information - which is something I hope to do better at in any software I may produce. I think the game Exapunks, while the system it teaches is a fictional system made up for the game, is a good example of a fairly high flow + high interactivity way of teaching skills.

I also think of the edutainment games I played as a kid, it's hard for me to highlight which ones I think are particularly good, since I haven't used them in a very long time, but I know they did a good job of using interactivity to force me to understand the concepts they taught. And I played them voluntarily, so they must have had at least decent flow.

Comment by mikkel-wilson on If there were an interactive software teaching Yudkowskian rationality, what concepts would you want to see it teach? · 2020-09-04T16:36:08.235Z · score: 1 (1 votes) · LW · GW

I think one crux between us is the degree to which "memory is the foundation of cognition", as Michael Nielsen once put it. Coming from the perspective that this is true, it seems to me that a natural consequence of a person memorizing even a simple sentence, and maintaining that memory with SRS, is that the sentence needs to be compressed in the mind to ensure that it has high stability, and can be recalled even after having not been used for many months, or even years.

In order to achieve this compression, it is inevitable that the ideas represented by the sentence will become internalized and integrated deeply with other parts of the mind, which is exactly what is desired. This process is a fundamental part of how the human mind works, and applies even in the mind of a person with low "rationality".

Comment by mikkel-wilson on If there were an interactive software teaching Yudkowskian rationality, what concepts would you want to see it teach? · 2020-09-03T17:54:18.734Z · score: 1 (1 votes) · LW · GW

While I agree with you that face to face interaction with a skilled mentor is the most effective way to learn complex skills such as rationality, that will fundamentally always be limited by the supply of humans who are both sufficiently skilled in the art, and are sufficiently good teachers, and who also have nothing better to do with their time.

So we really shouldn't look at this as either/or - we should, on the one hand, make sure there's good availability and supply of the best opportunity possible (face-to-face with skilled mentors), but also for the vast majority of learners for whom it isn't feasible to provide skilled human guidance, we need to provide the highest-quality content that can easily be scaled. There are flaws I see in the current best scalable solution (primarily stemming from a lack of interactivity), and I'm currently in a better position to attempt to address that issue than to improve the availability of human mentors

Comment by mikkel-wilson on Ideas for an action coordination website · 2020-09-02T17:44:03.742Z · score: 1 (1 votes) · LW · GW

Regarding communities: I don't think it's necessary for such a website to have a particularly strong conception of communities, outside of each particular collective action. For each collective action, there should be robust tools to make it clear who is / isn't qualified to take part in that action, but this should be handled on an action-by-action basis, and there's not much benefit to having a community structure outside of individual actions. Such a structure would simply add needless effort to the development and maintenance of such a website, which will make it less likely to achieve its mission in the long run.

Comment by mikkel-wilson on MikkW's Shortform · 2020-09-02T17:25:27.513Z · score: 3 (2 votes) · LW · GW

The phrase "heat death of the universe" refers to two different, mutually exclusive possibilities:

  1. The universe gets so hot, that it's practically impossible for any organism to maintain enough organization to be able to sustain itself and create copies of itself Or:
  2. The universe gets so cold, that everything freezes to death, and no organism can put make work happen to create more copies of itself

Originally, the heat death hypothesis referred to #1, we thought that the universe would get extremely hot. After all, heat death is a natural consequence of the second law of thermodynamics, which states that entropy can only increase, never decrease, and ceteris paribus (all else equal) when entropy increases, temperature also increases.

But ceteris is never actually paribus, and in this case, physicists found out that the universe is constantly getting bigger, things are always getting further apart. When volume is increasing, things can get colder even as entropy increases, and physicists now expect that, given our current understanding of how the universe works, possibility #2 is more likely, the universe will eventually freeze to death.

But our current understanding is only ever the best guess we can make of what the laws of the universe actually are, not the actual laws themselves. We currently expect the universe will freeze, but we could very well find evidence in the future that the universe will burn instead. Maybe (quite unlikely) things will just happen to balance out, so that the increase in temperature due to entropy equals the decrease in temperature due to the expansion of the universe.

Perhaps we will discover a loophole in a set of laws that would otherwise suggest a heat death of one kind or the other, but where a sufficiently intelligent process can influence the evolution of temperature so as to counteract the otherwise prevailing temperature trend - in the vein of (I'd like to note that I do not intend to imply that any of these are likely to happen) creating a large enough amount of entropy to create a permanent warm zone in a universe that is otherwise doomed to freeze (this would probably require a violation of the conservation of energy that we currently have no reason to believe exists), or using an as-yet undiscovered mechanism to accelerate the expansion of the universe that can create a long-lasting cool zone in a universe that is otherwise doomed to burn.

Comment by mikkel-wilson on moridinamael's Shortform · 2020-09-01T23:15:46.843Z · score: 1 (3 votes) · LW · GW
It's pretty unlikely that two things with independent 30% odds of being true, are both true.

I'm not sure I'd call 9% (the combined probability of two independent 30% events) "pretty unlikely" - sure, it won't happen in most cases, but out of every 11 similar situations, you would see it happen once, which adds up to plenty of 9% chance events happening all the time

Comment by mikkel-wilson on MikkW's Shortform · 2020-09-01T20:08:44.959Z · score: 1 (1 votes) · LW · GW
Why do you care about "most populous" or "densest (population over area)", regardless of definition of location?

1) Population density has an important impact on the mileau and opportunities that exist in a given location, but we can only make meaningful comparisons when metrics are standardized. 2) I've heard it said that in medieval times, many lords would collect a "bushel" of taxes from the peasants, where the bushel was measured in a large basket, but then when paying a "bushel" of taxes to their king, the bushel would be measured with a much smaller basket, thereby allowing the lord to keep a larger amount of grain for himself. When we don't have consistent standards for metrics, similar failure modes can arise in (subtler) ways - hence why I find reliance on arbitrary definitions of location to have bad taste

Comment by mikkel-wilson on MikkW's Shortform · 2020-08-31T02:50:52.880Z · score: 4 (3 votes) · LW · GW

There's plenty of land area for solar. I did a rough calculation once, and my estimate was that it'd take roughly twice the land area of the Benelux to build a solar farm that produced as much energy per annum as the entirety of humanity uses each year (The sun outputs an insane amount of power, and if one steps back to think about it, almost every single joule of energy we've used came indirectly through the sun - often through quite inefficient routes). I didn't take into account day/night cycles, or losses of efficiency due to transmission, but if we assume 4x loss due to nighttime (probably a pessimistic estimate) and 5x loss due to transmission (again, being pessimistic), it still comes out to substantially less than the land we have available to us (About 1/3 the size of the Sahara desert)

Comment by mikkel-wilson on MikkW's Shortform · 2020-08-31T02:44:50.251Z · score: 3 (2 votes) · LW · GW

The way we measure most populous cities / most dense cities is weird, and hinges on arbritary factors (take, for example, Chongqing, the "most populous city", which is mostly rural land, in a "city" the size of Austria)

I think a good metric that captures the population / density of a city is the number of people that can be reached with half an hour's or an hour's worth of transportation (1/2 hour down and 1/2 hour back is one hour both ways, a very common commute time, though a radius of 1 hour each way still contributes to the connections available) - this does have the effect of counting a larger area for areas with better transportation, but I think that's a good feature of such a metric.

This metric would remove any arbitrary influences caused by arbitrary boundaries, which is needed for good, meaningful comparisons. I would very much like to see a list organized by this metric.

(Edited: misremembered commute times. See Anthropological Invariants in Travel Behaviour)

Comment by mikkel-wilson on MikkW's Shortform · 2020-08-31T01:06:28.331Z · score: 1 (1 votes) · LW · GW

The big advantage chlorophyll has is that it is much cheaper than photovoltaics, which is why I was saying (in our conversation) we should take inspiration from plants

Comment by mikkel-wilson on MikkW's Shortform · 2020-08-31T00:47:42.090Z · score: 9 (5 votes) · LW · GW

During today's LW event, I chatted with Ruby and Raemon (seperately) about the comparison between human-made photovoltaic systems (i.e. solar panels), and plant-produced chlorophyll. I mentioned that in many ways chlorophyll is inferior to solar panels - consumer grade solar panels operate in the 10% to 20% efficiency range (i.e. for every 100 joules of light energy, 10 - 20 joules are converted into usable energy), while chlorophyll is around 9% efficient, and modern cutting edge solar panels can go even as high as nearly 50% efficiency. Furthermore, every fall the leaves turn red and fall down to the ground only for new leaves – that is plant-based solar panels – to be generated again in the spring. One sees green plants where there very well could be solar panels capturing light, and naïvely we would expect solar panels to do a better job, but we plant plants instead, and let them gather energy for us.

One of them (I think Ruby) didn't seem convinced that it was fair to compare solar panels with chlorophyll – is it really an apples to apples comparison? I think it is a fair comparison. It is true that plants do a lot of work beyond simply capturing light, and electricity goes to different things than what plants do, but ultimately what both plant-based farms and photovoltaic cells do is they capture energy from sunlight coming to the earth from the sun, and convert them to human usable energy. One could imagine genetically engineered plants doing much of what we use electricity for these days, or industrial processes being hooked up to solar panels that do the things plants do, and in this way we can make a meaningful comparison of how much energy plants allow us to use for human desired goals and compare that to how much energy photovoltaic cells can redirect to human-desired uses.

Comment by mikkel-wilson on How hard would it be to change GPT-3 in a way that allows audio? · 2020-08-28T15:05:22.989Z · score: 3 (2 votes) · LW · GW

I presume you mean GPT-3?

Comment by mikkel-wilson on Spaced Repetition Database for A Human's Guide to Words · 2020-08-24T14:57:40.468Z · score: 1 (1 votes) · LW · GW

Some more thoughts: there's one card in my notes that asks:

"According to Ray Dalio, when you have a problem, you should ask yourself 6 questions: What went wrong? Have you made a mistake like this before? What was the immediate cause of the problem? What was the root cause of the problem? What can you do to correct the problem in the short term? What can you do to prevent problems like these in the long term?"

All 6 questions are hidden on the same card, so I have to provide all 6 in order to mark the card correct. Following the principle of atomization, I might have thought to create 6 different cards, one for each question to ask, but my experience with such atomized lists tends to be that each card is much more slippery, and the different items would all blend together in my mind (creating problems both during review and when the opportunity for application arises), whereas when the list is presented as a cohesive whole, it's much easier to remember each part as a part of the whole, and makes each part contrast better with the other items

Comment by mikkel-wilson on Inoculating against Psychedelic Woo · 2020-08-24T05:50:30.215Z · score: 1 (1 votes) · LW · GW

I don't think Nootropics is the right tag for the concept I want to index- if I were ever to go to the Nootropics tags, I'd expect to find conversation about substances I could generally expect to take on a regular basis to improve my cognition. While I think the case can be made for microdosing being nootropic, microdosing is only mentioned here in the context of having a pseudo-trip experience, which is different in intent from nootropic microdosing.

The concept I'm looking to index is substances that create short-term experiences which lead to long-term influences on one's life, which is an overall different concept from nootropics

Comment by mikkel-wilson on Sunday August 23rd, 12pm (PDT) – Double Crux with Buck Shlegeris and Oliver Habryka on Slow vs. Fast AI Takeoff · 2020-08-24T04:27:46.798Z · score: 16 (5 votes) · LW · GW

I just wanted to register my enjoyment with the hangout we all had in Topia post-talk, and I look forward to such hangouts in the future, and hope they continue even when things return to normal

(I'll also mention that I vastly preferred the Topia experience to Zoom hangouts I've attended in the past, though I was frustrated that Topia forced me to use my desktop computer and turn off my iPad, which I usually use for chats)

Comment by mikkel-wilson on [Geo: California/Bay Area] Mini-Research of 2020 Fire Situation · 2020-08-24T03:19:36.453Z · score: 5 (3 votes) · LW · GW

Worth noting, there's going to be more lightning tonight, which may or may not spark more fires

Comment by mikkel-wilson on Spaced Repetition Database for the Mysterious Answers to Mysterious Questions Sequence · 2020-08-21T23:49:58.264Z · score: 1 (1 votes) · LW · GW

You should be able to do a "custom review", which allows you to specify which cards to review, and you can tell it to show you cards from multiple decks

Comment by mikkel-wilson on Spaced Repetition Database for A Human's Guide to Words · 2020-08-21T23:37:44.189Z · score: 3 (2 votes) · LW · GW
Normal flashcards should be all equally difficult: as easy as possible. The idea is to break everything down into atomic facts; this makes it so you can't short-circuit a difficult card by just memorizing the answer; by memorizing all the parts, you still have the whole.

In my experience, I've found this to not be as true as it seems. I originally had many of my cards as atomic as possible, based on what Piotr Wozniak suggests, but while I had each individual part memorized, they all seemed to hang loosely connected, without a vivid thread holding it all together. Two clear examples for me are memorizing chess moves and poetry: Originally I'd memorize poems line by line, with each line shown in its surrounding context to prompt me, and chess games move-by-move, prompted by the state of the board at that moment.

I later experimented with more coarse chunks, where a single card would represent an entire stanza of a poem, or 10 ply of moves from a high-tier chess game. The cards took longer per interaction, but only a little longer (since I didn't have to switch contexts, increasing short-term flow), and the relativly fewer number of cards needed to represent the information more than made up for the longer reviews. But the most important benefit was that things suddenly became much clearer for me. Instead of being vaguely aware of the contours of the poems, I would find myself reciting them wholesale in the shower, unprompted (even fairly long poems like Tennyson's Ulysses, which clocks in at 70 lines - which, by the way, I shudder at the thought of trying to memorize line-by-line, like I have some shorter poems). Suddenly, I would see an entire series of chess moves as being deeply interconnected, in a way I never noticed trying to memorize those same moves one at a time.

The main benefit I thought atomization would provide, namely that each element would be readily available for combination with other, seemingly unrelated ideas, I don't really notice much loss of. While each element is richly situated in an existing context, I can easily pluck that idea out of that context to join it with ideas from another, but the richness of the existing context makes it much easier to get a handle on that idea in the first place

Comment by mikkel-wilson on Inoculating against Psychedelic Woo · 2020-08-21T23:00:34.799Z · score: 1 (1 votes) · LW · GW

Strong upvoted. That's an important warning to keep in mind

Comment by mikkel-wilson on Inoculating against Psychedelic Woo · 2020-08-21T21:46:59.506Z · score: 3 (2 votes) · LW · GW

On reflection, the method I described seems very cautious, perhaps overly cautious in its approach. Honestly, I think putting aside what I wrote above, my experiences with other (legal) mind-influencing substances has been much more in the style of charging forward, and just experiencing the effect a normal-sized dosage provides and only later, when I'm sober, reflecting on the effect the substance has on me.

So, the exact procedure I describe above may very well not be the way I'd approach psychedelics, although I think perhaps some part or another may provide useful inspiration in engineering a better experience

I'm also not sure how much of a problem "psychedelic woo" may be for someone already well rooted with epistemic defenses- perhaps the things that lead people astray can be untangled by a sufficiently careful thinker experiencing them, especially if psilocybin / peyote is used, as opposed to LSD, which supposedly has stronger negative effects on the mind. But I'm also pessimistic about the average LWer's epistemic defenses- simply reading The Sequences, without doing any exercises based on the ideas presented therein, won't give people the skills needed to successfully navigate the challenge of properly making sense of a psychedelic experience. It is a well known phenomenon in learning, that simply reading isn't enough. One needs to interact with the ideas, be rewarded for demonstrating understanding, and having their lack of understanding highlighted where it's missing. That's how people form strong connections to their ideas (Anki would be sufficient for this, based on my own experiences with Anki, although there may be even better ways of accomplishing this, at the cost of higher effort needed in implementation).

I think the version of my presented idea, which I still endorse, is that before experiencing psychedelics, if one wishes to gain the benefits, while avoiding well-known long-term pitfalls, one should strive to have a robust understanding of epistemic defenses - to have actually interacted with the ideas, felt the ideas push back against them, not just have read about them. This is not to (as my post above suggests) prevent irrational thought during the trip itself, but rather to ensure that one is properly prepared, once sober, to reflect critically about their experiences, and avoid common epistemic mistakes people make post-psychedelics

Comment by mikkel-wilson on Inoculating against Psychedelic Woo · 2020-08-21T21:32:45.378Z · score: 2 (2 votes) · LW · GW

It's a good idea. I'm not familiar with any existing decks, but a search on AnkiWeb shows a few LW-influenced decks: https://ankiweb.net/shared/decks/Rationality. One of them looks like what I would be inclined to design as an "Epistemic Defenses Deck", although I may (or may not) approach it differently. If I do anything along those lines, I'll let you know.

Comment by mikkel-wilson on ricraz's Shortform · 2020-08-21T15:25:00.554Z · score: 3 (2 votes) · LW · GW

AI Alignment Forum karma (which is also displayed here on posts that are crossposted)

Comment by mikkel-wilson on Calibrate words, not just probabilities · 2020-08-21T06:18:23.330Z · score: 1 (1 votes) · LW · GW

I mostly agree with this comment. I do think there are broad categories we can put words / phrases (and probably body language and other paralinguistics) into, which can give us meaningful evidence of the other person's confidence.

Comment by mikkel-wilson on Inoculating against Psychedelic Woo · 2020-08-21T06:04:33.907Z · score: 5 (3 votes) · LW · GW

I'm trying to think of the right tag for this post- It seems that it'd be mildly relevant for there to be a tag for psychedelics and other mind-altering substances, but I'm not sure what the right name for the tag would be. "Drugs" feels too broad, "Psychedelics" is too narrow, since there are substances which are technically not psychedelics, that can serve similar functions of having short-term effects on the brain's biology, but long-term (positive or net-neutral) effects on the brain's nature and structure. "Mind-altering substances" seems right, but I'm not sure that there isn't a better way to put it

Comment by mikkel-wilson on Viliam's Shortform · 2020-08-21T05:23:55.029Z · score: 1 (1 votes) · LW · GW

Ah, good ol' Freigeld

Comment by mikkel-wilson on MikkW's Shortform · 2020-08-20T18:06:45.390Z · score: 1 (1 votes) · LW · GW

As an aside, I think it's worth pointing out that Esperanto's use of the prefix mal- to indicate the opposite of something (akin to Newspeak's un-) is problematic: two words that mean the exact opposite will sound very similar, and in an environment where there's noise, the meaning of a sentence can change drastically based on a few lost bits of information, plus it also slows down communication unnecessarily.

In my notes, I once had the idea of a "phonetic inverse": according to simple, well defined rules, each word could be transformed into an opposite word, which sounds as different as possible from the original word, and has the opposite meaning. That rule was intended for an engineered language akin to Sona, so the rules would need to be worked a bit to have something good and similar for English, but I prefer such a system to Esperanto's inversion rules

Comment by mikkel-wilson on Viliam's Shortform · 2020-08-20T04:43:49.447Z · score: 3 (2 votes) · LW · GW

I would very much like to see a society where money circulates very quickly. I expect people will have many reasons to be happier and suffer less than they do now.

As you observe, income taxes encourage slowing down circulation of money, while wealth taxes speed up circulation of money (and creation of value), but I think there are better ways of assessing tax than those two. I suspect heavily taxing luxury goods which serve no functional purpose, other than to signal wealth, is a good direction to shift taxes towards, although there may be better ways I haven't thought of yet.

Not answering your question, just some thoughts based on your post

Comment by mikkel-wilson on Vanessa Kosoy's Shortform · 2020-08-20T03:13:21.731Z · score: 1 (1 votes) · LW · GW

I have a sense that similar principles are at play with Spaced Repetition, and that pointing out that connection may be relevant to effectively handling this issue

Comment by mikkel-wilson on For the past, in some ways only, we are moral degenerates · 2020-08-19T21:00:37.270Z · score: 1 (1 votes) · LW · GW

I certainly agree that we'd probably have a hard time convincing someone from the past to be able to understand that.

I hope for myself, and my cohort generally, that we'd be able to rise above the level of being attached to something that doesn't actually matter, and sacrifice superficial values for what we care most about

Comment by mikkel-wilson on For the past, in some ways only, we are moral degenerates · 2020-08-18T23:35:26.278Z · score: 3 (2 votes) · LW · GW

It seems to me that in the cases of extended family values and {honor and reputation}, both modern values and past values are indirect attempts to optimize the same hidden objective. While a modern person may be shocked by past values in these categories, and a past person shocked by modern values, if we digged deep enough into why we actually cared about these values, we would find that we care for them for the same reasons, and measurable progress can be made on that shared base-level value, which is more important than the surface-level values which may be change over time

Comment by mikkel-wilson on Why haven't we celebrated any major achievements lately? · 2020-08-18T15:22:03.736Z · score: 2 (2 votes) · LW · GW

I think while there are many who feel uncomfortable showing national pride specifically, there's still capacity for collective pride- just a different collective, not the nation