Jackercrack

Karma: 159

Jackercrack Nov 1, 2014, 10:17 AM
0 points
in reply to: the-citizen’s comment on: Open thread, Oct. 20 - Oct. 26, 2014
Agreed, a general rule of truthiness is definitely a very effective approach and probably the most effective approach, especially once you’ve started down the path. So far as I can tell stopping halfway through is… risky in a way that never having started is not. I only recently finished the sequences myself (apart from the last half of QM). At the time of starting I thought it was essentially the age old trade off between knowledge and happy ignorance, but it appears at some point of reading the stuff I hit critical mass and now I’m starting to see how I could use knowledge to have more happiness than if I was ignorant, which I wasn’t expecting at all. Which sequences are you starting with?

By the way, I just noticed I screwed up on the survey results: I read the standard deviation as the range. IQ should be mean 138.2 with SD 13.6, implying 95% are above 111 and 99% above 103.5. It changes my first argument a little, but I think the main core is still sound.

Jackercrack Nov 1, 2014, 1:42 AM
0 points
in reply to: Lumifer’s comment on: A discussion of heroic responsibility
Persistence is a good word for it, plus a sense of making it work even if the world is unfair, the odds are stacked against you. No sense of having fought the good fight and lost, if you failed and there were things you possibly could done beforehand, general strategies that would have been effective even if you did not know what was coming, then that is your own responsibility. It is not, I think, a particularly healthy way of looking at most things. It can only really be useful as a mindset for things that really matter.

can you be more explicit?

Ah, sorry, I insufficiently unpacked “effective ways to satisfy terminal values”. The hidden complexity was in “effectively”. By effectively I meant in an efficient and >75% optimal manner. Many people do not know their own terminal values. Most people also don’t know that what makes a human happy, which is often different from what a human wants. Of those that do know their values, few have effective plans to satisfy them. Looking back on it now, this is quite a large inferential distance behind the innocuous looking work ‘sane’. I shall try to improve on that in the future.

Jackercrack Nov 1, 2014, 12:10 AM
0 points
in reply to: Lumifer’s comment on: A discussion of heroic responsibility
Then what was all that stuff on the news about cutting government jobs, trying desperately to ensure frontline services weren’t effected and so on about?

Edit: I knew it! No wonder I felt so confused. It would seem the reduction in spending just took a while to come into effect. Take a look at the years after 2011 that your chart is missing. Unfortunately it’s not adjusted for inflation but you still get the idea. If you change category to protection and the subcategory to ‘police’, ‘prisons’ or ‘law courts’, you can see the reduction in police funding over the course of the recession.

Jackercrack Oct 31, 2014, 9:14 PM
0 points
in reply to: Lumifer’s comment on: A discussion of heroic responsibility
There does come a point when the bill must be paid though, even if it is over a long time. Even if it’s over 40 years as you pay back the interest on the debt.

Before we go further, I think we need to be sure we’re talking about the same thing when we say power. See, when you said a reduction in government power, what I heard was essentially less money, smaller government. I’m getting the feeling that that is not entirely what you meant, could you clarify?

Jackercrack Oct 31, 2014, 9:10 PM
0 points
in reply to: Lumifer’s comment on: A discussion of heroic responsibility
Well, yes, it was all over the news. This feels like a trick question. Are you about to tell me that spending went up during the recession or something?

Jackercrack Oct 31, 2014, 5:52 PM
0 points
in reply to: V_V’s comment on: A discussion of heroic responsibility
They don’t necessarily have to, but generally do. For instance during austerity measures spending is generally reduced in most areas. Police forces have less funding and thus lose the ability to have as great an effect on an area, that is they have less power. Unless you’re talking about power as a state of laws instead of a state of what is physically done to people?

Jackercrack Oct 31, 2014, 5:20 PM
0 points
in reply to: Nornagest’s comment on: A discussion of heroic responsibility
Yes, I was going to comment on how a clever politician could spend during their own term to intentionally screw over the next party to take power, but I wanted to avoid the possible political argument that could ensue.

Jackercrack Oct 31, 2014, 4:32 PM
0 points
in reply to: Lumifer’s comment on: A discussion of heroic responsibility
All of it looks reasonable to me apart from the last paragraph. I can see times when governments do willingly contract. There are often candidates who campaign on a platform of tax cuts, the UK had one in power from 1979-1990 and the US had one in power from 2001-2009.

Tax cuts necessarily require eventual reductions in government spending and thus the power of government, agreed?

Jackercrack Oct 31, 2014, 10:18 AM
2 points
in reply to: Lumifer’s comment on: A discussion of heroic responsibility
I’ll put this in a separate post because it is not to do with heroic responsibility and it has been bugging me. What evidence do you have that your favoured idea of reducing political power does what you want it to do? Are there states which have switched to this method and benefited? Are there countries that have done this and what happened to them? Why do you believe what you believe?

Jackercrack Oct 31, 2014, 10:14 AM
0 points
in reply to: Lumifer’s comment on: A discussion of heroic responsibility
Okay, my definition of sane is essentially: rational enough to take actions that generally work towards your goals and to create goals that are effective ways to satisfy your terminal values. It’s a rather high bar. Suicide bombers do not achieve their goals, cultists have had their cognitive machinery hijacked to serve someone else’s goals instead of their own. The reason I think this would be okay in aggregate is the psychological unity of mankind: we’re mostly pretty similar and there are remarkably low numbers of evil mutants. Being pretty similar, most people’s goals would be acceptable to me. I disagree with some things China does for example, but I find their overwhelming competence makes up for it in aggregate wellbeing of their populace.

gjm gives some good examples of heroic responsibility, but I understand the term slightly differently. Heroic responsibility is to have found a thing that you have decided is important, generally by reasoned cost/benefit and then take responsibility to get it done regardless of what life throws your way. It may be an easy task or a hard task, but it must be an important task. The basic idea is that you don’t stop when you feel like you tried, if your first attempt doesn’t work you do more research and come up with a new strategy. If your second plan doesn’t work because of unfair forces you take those unfair forces into account and come up with another plan. If that still doesn’t work you try harder again, then you keep going until you either achieve the goal, it becomes clear that you cannot achieve the goal or the amount of effort you would have to put into the problem becomes significantly greater than the size of the benefit you expect.

For example, the benefit for FAI is humanities continued existence, there is essentially no amount of effort one person could put in that could be too much. To use the example of Eliezer in this thread, the benefit of a person being happier and more effective for months each year is also large, much larger than the time it takes to research SAD and come up with some creative solutions.

Jackercrack Oct 31, 2014, 1:18 AM
2 points
in reply to: Lumifer’s comment on: A discussion of heroic responsibility
See, you’re ignoring the qualifier ‘sane’ again. I do not consider suicide bombers sane. Suicide bombers are extreme outliers, and they kill negligible numbers of people. Last time I checked they kill less people per year on average than diseases I had never heard of. Quite frankly, they are a non-issue when you actually look at the numbers.

It is not obvious to me that heroic responsibility implies that a thing should be done without cost/benefit analysis or at any cost.

Of course it depends on the values systems involved, I just happen to be fine with most values systems. I’ll rephrase normal values systems to be more clear: People will on average end up with an average range of value systems. The majority will probably be somewhat acceptable to me, so in aggregate I’m fine with it.

Is there a specific mechanism by which reducing government power would do good? What countries have been improved when that path has been taken? It seems like it would just shift power to even less accountable companies.

Jackercrack Oct 30, 2014, 11:49 PM
0 points
in reply to: Lumifer’s comment on: A discussion of heroic responsibility
I think we’re using a different meaning of the word sane. See, I hold sanity to a rather high standard which excludes a huge breadth of people, probably myself as well until I’ve progressed somewhat.

I am imagining enough sane people taking heroic responsibility, the world looks rather different than this and it seems to be better run. We already have people in charge with value systems unacceptable to me, making them at least competent and getting them to use evidence-based strategies seems like a step forwards. People will have a normal range of value systems, if a particularly aberrant person comes with a particularly strange value system, then they’ll still have to outsmart all the other people to actually get their unacceptable value system in place

Honestly lumifer, I’m beginning to thing you never want to change anything about any power structure in case it goes horribly wrong. How are things to progress if no changes are allowed?

Jackercrack Oct 30, 2014, 11:24 AM
4 points
on: Meetup : Bath, UK: Agreement, practical meetups, and report from last meetup
See you on sunday! I’ve had a look at some cool exercises to do and Fermi Calculations, 5 minute debiasing and Zendo all look fun and useful. We’ll talk it over it at the rock climbing place in a few hours

Jackercrack Oct 30, 2014, 8:34 AM
0 points
in reply to: Lumifer’s comment on: A discussion of heroic responsibility
You say he’s not-mad, but isn’t he the spitting image of the revolutionary that power corrupts? Wasn’t Communism the archetype of the affective death spiral?It would appear he was likely suffering from syphilis, a disease that can cause confusion, dementia and memory problems. Anyway, isn’t that an ad hominem argument?

Jackercrack Oct 29, 2014, 11:51 PM
1 point
in reply to: ChristianKl’s comment on: Open thread, Oct. 27 - Nov. 2, 2014
Any idea how long that will be (months, years, decades)?

Jackercrack Oct 29, 2014, 11:37 PM
0 points
on: A discussion of heroic responsibility
I think heroic responsibility is essentially a response to being in a situation where not enough people are both competent at and willing to make changes to improve things. The authority figures are mad or untrustworthy, so a person has to figure out their own way to make the right things happen and put effective methods in place. It is particularly true of HPMOR where Harry plays the role of Only Sane Man. So far as I can tell, we’re in a similar situation in real life at the minute: we have insufficient highly sane people taking heroic responsibility. If we had enough sane people taking heroic responsibility, things would look rather different and likely a lot better run. It would be easy to be a happy gear, if you knew the machine was properly designed, the goal was a good one and the plan was likely to succeed.

There is clearly some ratio of heroically responsible:useful gears that works the best for each situation, some optimal equilibrium. Too many people trying to unilaterally change things in different directions and you get chaos and infighting. Too many useful gears and you have a wonderfully maintained, smooth running machine working at 40% maximum efficiency towards a goal that doesn’t make much sense. I propose heroic responsibility be fit into a larger framework, that of filling the role that is required. I can’t think up a snappy name for it, but you essentially mould your actions into the shape that best maximises your group’s outcome given your abilities. If there is already someone doing the job of heroic responsibility and doing it well, you aim for the next empty or poorly-done role down where you do most good. If all the important positions are filled by competent people or by people more competent than you, then be a gear.

The problem lies in knowing whether you actually could do better in someone’s situation with the resources available to them. The ethical injunctions seem to say that humans are rather bad at figuring out when they could do better than those in power. There is only an injunction against cheating to gain power though, not against legitimate means. Still, it’s difficult to say whether one is actually more competent and a large amount of research into the potential role should be done first before action is taken. If you can see a decision for the role coming and make a strong prediction of what a good choice would look like, then seen what the actual decision the person made was and predict the effects and finally get it right, with a better batting average than the person in the role, if you can do that, then it’s time to go for the position.

Disclaimer: There is a possibility that this theory is an extension of my belief that I should eventually be in charge.

Jackercrack Oct 29, 2014, 5:46 PM
1 point
in reply to: the-citizen’s comment on: Open thread, Oct. 20 - Oct. 26, 2014
I’m not sure they’re wrong to be honest (assuming an average cross section of people). Rationality is an extremely long term approach and payoff, I am not sure it would even work for the majority of people and if it does I’m not sure if it reaches diminishing returns compared to other strategies. The introductory text (sequences) is 9,000 pages long and the supplementary texts (kahneman, ariely ect) take it up to 11,000. I’m considered a very fast reader and it took me 3 unemployed months of constant reading to get through. For a good period of that time I was getting a negative return, I became a worse person. It took a month after that to end up net positive. I don’t want to harp on about unfair inherent advantages, but I just took a look at the survey results from last year and the lowest IQ was 124.6. This stuff could be totally ineffective for average people and we would have no way of knowing. Simply being told the best path for self improvement or effective action by someone who was a rationalist or just someone who knows what they’re doing, a normal expert in whatever field may well be more optimal for a great many people. Essentially data-driven life coaching. I can’t test this hypothesis one way or the other without attempting to teach an average person rationalism and I don’t know if anyone has done that, nor how I would find out if they had.

So far as instrumental rationality not being in core about truth, to be honest I broadly agree with them. There may be a term in my utility function for truth but it is not a large term, not nearly so important as the term for helping humanity or the one for interesting diversions. I seek truth not as an end in itself, but because it is so damn useful for achieving other things I care about. If I were in a world where my ignorance would save a life with no downside while my knowledge had no longterm benefit then I would stay ignorant. If my ignorance was a large enough net benefit to me and others, I would keep it. In the arena of CEO compensation for example increased transparency leads to runaway competition between CEOs to have the highest salary, shafting everyone else. Sure, the truth is known but it has only made things worse. I’m fairly consequentialist like that.

Note that in this situation I’d still call for transparency on war crimes, torture and so on. The earlier the better. If a person knows that their actions will become known within 5 years and that it will effect them personally that somewhat constrains their actions against committing an atrocity. The people making the decisions obviously need accurate data to make said decisions with in all cases but the good or damage caused by the public availability of that data is another thing entirely. Living in a world where everyone was rationalists and the truth never caused problems would be nice, but that’s the should-world not the is-world.

It so happens that in this world we live using these brains we have, seeking the truth and not being satisfied with a lie or a distortion is an extremely effective way to gain power over the world. With our current hardware truth seeking may be the best way to understand enough to get things done without self-deception, but seeking the truth itself is incidental to the real goal.

Jackercrack Oct 29, 2014, 6:11 AM
2 points
in reply to: gjm’s comment on: Open thread, Oct. 27 - Nov. 2, 2014
I rather think there may be demand for a cheaper, less time dependent method of attending. It may be several seasons before they end up back in my country for example. Streaming/recording the whole thing and selling the video package seems like it could still get a lot of the benefits across. Their current strategy only really makes sense to me if they’re still in the testing and refining stage.

Jackercrack 29 Oct 2014 6:05 UTC
2 points
in reply to: cursed’s comment on: Open thread, Oct. 27 - Nov. 2, 2014
Well that’s a bit dispiriting, though I suppose looking back my view of CFAR was a bit unrealistic. Downregulating chance that CFAR is some kind of panacea.

Jackercrack 29 Oct 2014 5:45 UTC
1 point
on: How to Run a Successful Less Wrong Meetup
Fantastic guide, a small note: the various status cues on page 28 is a dead link

Edit: Note to future readers: dead links can generally be traversed using the internet archive to find the cached site from when it was up.