Author Archives: david

The Other Hammer Principle

If you google for “hammer principle” you get two things: The site mike and I created, and a bunch of stuff about martial arts.

Well I’ve been thinking about learning a martial art, so they were on my mind, and it occurred to me that martial arts were pretty ideal for the sort of thing that hammer principle does.
.
So we created a Martial Arts Hammer Principle. It’s pretty incomplete so far – only 14 responses – but already it’s starting to take shape as some quite interesting results.

So if you know any martial arts, or are interested in learning one, I encourage you to head over and check it out! And if you can fill out some of the statements and help us build a picture of different martial arts, that would be even better.

This entry was posted in Uncategorized on by .

World, whim, etc. Part 2

As has been observed, my The World As I Would Make It leaves a lot of policy details undefined. That was deliberate – some I haven’t put enough thought into or acquired enough knowledge to define. Some I just didn’t want to include because it would make the post too sprawling.

One issue I care a lot about but have not yet put enough study (and, to a lesser extent, thought) into is the structure of the government itself. I thought I’d sketch something out anyway, but I’m not at all convinced that what I’m describing here is a great idea. I like the overall shape of it and the aims are valid, but a lot of the individual details may be suspect. In particular, any numbers present here are basically made up and should be regarded as complete wild assed guesses.

Basic Government Structure

The structure of the government is similar to the classic two house system as seen in e.g. the UK. You’ve got the House, a body of 100 candidates whose role it is to pass laws (analogous to the house of commons), and you’ve got Oversight (analogous to the house of lords) who are required to approve it.

Additionally you have a prime minister (and maybe deputy prime minister?) who is not a member of either body (and does not get a vote in them) who acts as head of government.

The prime minister is elected during the general election, which is not tied to the appointment of members of either house (more on why later). This may be called for by a majority vote in any one of House, Oversight, or the electorate (or the existing prime minister may call for it. However this doesn’t differ much from explicitly stepping down).

Voting

The system is, strictly speaking, not a democracy. It’s a meritocracy. However the conditions for merit are not a very high bar to meet – they are designed to ensure a basic level of competence and interest.

The reason for this a lot of you will disagree with me on: I don’t regard voting as a right. The role of a government is not to obey the will of the people, it’s to maximize the benefit to the people. Citizen oversight is essential for this, and democracy is currently one of the best ways we have of achieving this. However in order of decreasing importance it goes benefit, oversight, democracy.

It’s currently far too easy to vote. This means that the vote is very manipulable by spin and media campaigning. I’m not in favour of making it very difficult to vote, but I am in favour of dampening this effect.

How to earn the vote:

  1. A school education is required (remember: Part of the curriculum in my desired school system is “critical thinking”, which includes a bullshit detector)
  2. Regular community service (this idea thanks to Dave Stark)

Beyond these criteria there is exactly one crime for which you can have your right to vote permanently (unless you later manage to prove yourself innocent of course) removed from you: Interfering with the democratic process. This includes but is not limited to vote rigging, putting pressure on someone else for how they would vote and bribery.

The way the community service works is this: You don’t have to do a lot. The requirement is probably something like you have to have averaged X hours per month for the last 6 months (which could be done weekly or in one batch or whatever you like). This is assigned to you randomly – it’s not skills tested, it’s not in any way reflective of privilege. One way to do this would be to have it so every 10 hours of work you are given two random options and may choose between them. If you genuinely feel you are unable to perform either of them you may appeal, but this shouldn’t be a straightforward process.

Details need to be worked out in terms of “grace periods” in which you can make up missed community service, etc. It’s important to get right, but for now I’m just going to assume that something sensible can be worked out and then tinkered with when it’s tried out in practice. People who work in the military should probably be automatically counted as performing community service while on duty (but be required to do some while on leave). Additionally people who are required to spend a lot of time out of the country (ambassadors, people who work on oil rigs, etc) may get special exemption to perform extra community service whilst in the country.

As a side note, I think you need to be eligible to vote when the vote is declared, not when the vote is performed. No last minute “oh shit I care about that, I’d better do some community service in a hurry”.

Peoples’ eligibility to vote

So. You’ve earned the vote. How does voting work in practice?

First, how is Oversight elected? It’s not. It’s a jury service. The members of Oversight are randomly selected from the electorate. You are allowed to refuse election. Elections are for (staggered) fixed terms, and you may step down at any time.

For the House the voting is a little more traditional. As we’ve discovered recently, this is quite a controversial subject and I haven’t fully made up my mind on my preferred system, but for the moment I’m going to borrow/paraphrase one I quite like from the Swiss. It’s a modified form of range voting (I don’t actually especially like pure range voting, but do mostly like the following modification).

Every voter gets 10 votes. They are not required to use all of them. These votes may be assigned to candidates. A voter may assign up to two votes to any given candidate. Tally the number of votes each candidate get and select the top N.

This system is used to elect members of the House: You vote for the entire house of representatives (it’s not a constituency system, although one expects that most representatives will target their campaigning at least somewhat geographically) and the top 100 representatives get in. I’m unsure what mechanism is used to elect the prime minister. If there’s also a deputy prime minister then possibly the same system with the top two selected.

Another novelty to the system:

Firstly, the system is electronically counted. This means the votes are not anonymous. I don’t know much about this subject, but the security community at large have basically said they’re unable to come up with an anonymous electronic voting system that isn’t trivially compromised. They are private, but the information exists and is centrally stored to link a vote back to a voter.

Additionally, complete anonymized vote data is published, and subject to random checks for accuracy, so the counts can be replicated by anyone.

Why is such electronic counting necessary? Well, partly because you’re voting for the entire house of representatives, so a physical user interface is somewhat prohibitive and you need electronic mechanisms of voting (this isn’t really true: You could have a ledger mapping each candidate to an ID and write the ID on a piece of paper. If it proves prohibitively difficult to build safe voting mechanisms even if the count is electronic this could be done, but it’s a much larger amount of work for voters, sensitive to handwriting and harder to count).

But more importantly is the innovation I like most about this design (other than the prerequisites for voting): There are no elections. Voting is a continuous process. You cast your vote and it remains in force until you choose to change it or are no longer eligible for the vote. You will be periodically asked to confirm that it is your vote and ask whether you’d like to change it. If a politician behaves in a way you regard as inappropriate you may immediately change your vote. Your vote always matters, because it is always in force and always subject to your control – there is no temptation to rest on your laurels and go “Ok, I’ve voted. I don’t need to do my community service now”.

When a candidate enters the top 100 another one is automatically booted out (by construction). There is a grace period for them to wrap up affairs (and possibly campaign to regain their position during that time) – otherwise there’s far too much volatility in the lower end of the House – and then their replacement comes in.

I also rather like the idea of having league tables which you can see politicians and candidates moving up and down in them according to current popularity.

There are definitely some downsides to this: The biggest one is that a candidate has a really strong incentive to only make popular decisions (which tend to be short-term gains in nature). I don’t know what to do about that. I hope the fact that the populace is better informed and that it requires at least a modicum of work to change your vote will help offset this. Also if you’re already quite high in the ranking then it’s harder to lose the vote so you’ve got more flexibility to make long-term decisions.

Edit: On discussion with Dave, an alternative and less fickle system is to have fixed terms and run annual elections for a fraction of the electorate. e.g. have everyone on a 5 year term and every year hold a vote for 20 of the candidates each year. I am undecided how I feel about this system, but it does retain some of the advantages of the above while being possibly less chaotic.

Representatives and candidates

One of the major problems with current representation is that, well, you have to be fairly rich to do it. I think this needs to be discouraged.

I think the easiest way to fix this is probably to make it easier for people who are not ridiculously wealthy to run. As such I’d like to propose the following program:

Anyone who wishes to may apply for candidacy aid. They must have 1000 signatures supporting their candidacy. They will now be paid a reasonable wage (probably hovering somewhere around population median) to campaign as a candidate. They have one year, during which they must demonstrate they are actively campaigning, to achieve some number of votes (maybe 10k?). If they achieve that they may continue on candidacy aid for as long as their vote remains above that number (possibly with some fixed maximum time period, or requiring demonstrating an annual increase).

It’s probably not enough, but it should be a good start.

Conclusion

Like I said, this is probably poorly thought out, but it contains ideas I think are good and may well be refineable into something actually worthwhile. Thoughts?

This entry was posted in Uncategorized on by .

The World as I Would Make It

I commented on twitter that the results of my imposing my beliefs and whims on the political landscape would offend a bunch of people. This got me to thinking that it might be interesting to actually think about what I would do. And, well, if I find it interesting then clearly the target audience of my blog would too, right? :) (hint: I’ve always adopted a policy that this blog is about what I find interesting, and that if you guys agree that’s great but not that important).

Unfortunately, it turns out to be quite hard to write down as a plan. It forms a coherent whole in my head, but figuring out what order to put it down in and how to present it is difficult. This is my third attempt at doing so…

Warning: What follows may be ill-informed, poorly thought out or merely naive. Some of it I’ve thought of in detail, some of it I haven’t. All of it I am fully persuadable it’s a bad idea (although you’re less likely to persuade me that the things that lead me to it are invalid).

Accountability and Empiricism

One of the biggest things I want out of government is a logic strategy of proving that things work. Policies need to have goals which they can prove they met. Different things need to be tried with randomized A/B testing. There need to be people having oversight and investigative power over the whole government.

The measure of all ideas presented here is that they work. If they turn out not to work, they must be replaced by something else. In order to do this, there need to be people in government whose sole responsibility is to be constantly examining what the government is doing and call attention to where it doesn’t work and how it could be improved (and have the power to make sure they are listened to)

Money

Aspects of this plan are expensive. I’m currently unsure how expensive – I’d need to do a detailed budget, which is almost impossible without a great deal of expensive research and experimentation. Most of this plan I don’t think is outrageously expensive. The education plan definitely is a lot more expensive than our current one (or is it? Parts of it are self-sustaining, and it may result in many students being in the system for less time, which might offset its cost by increasing the capacity. Unlikely but possible). There area also things, like healthcare, where although the plan itself is not inherently expensive I consider it strongly desirable to spend more on them anyway.

There are places I believe you can save money in the current system – much of it is inefficient, and I dislike the amount of money spent on the military (If I could come up with a coherent plan for disbanding the military altogether I would. I’d also give everyone a free magical flying unicorn pony. But I do believe that you can have a realistic purely defensive military for a fraction of the cost of sustaining a modern western country’s military structure). So I honestly have no idea what government spending would be like if I had my way. However in the bits I actually intend to cover here, it’s definitely higher.

As a result, taxes are higher. Sorry. I’m ok with that – I think you get much much better results per tax-unit-of-currency than you do with the current system, and I’m ok with people paying more.

Tax is very much still banded, and the tax increase probably should affect the wealthy proportionately more. Fiscal conservatives, feel free to start yelling at me now and disregard the rest of this post.

Benefits and unemployment

Someone who isn’t me suggested this. I can no longer remember who – it was a blog post somewhere – but I thought it was an excellent idea.

There is no “dole”, no jobseekers’ allowance. No benefits paid to those unemployed.

There is however a “base living wage” which the government pays – it’s not a lot, it’s certainly not enough to live comfortably, but it is enough to live. This wage is untaxed and uniformly applied across the country.

It is additionally the government’s responsibility to build hostels. These are not shelters – they are paid accomodation – however the price is fixed at a level where someone on only the base wage can afford to live and eat there and have a little bit left over. The government is legally obliged to build hostels to meet capacity in the presence of overcrowding (and maybe to offer transport to other locations if people are unable to find a hostel near them with capacity).

The ideal here is to make sure that no one (who is behaving rationally, which I admit isn’t everyone)

  • is unable to find/afford somewhere to live
  • is given an incentive to not find a job (switching from the dole to a minimum wage job in the UK is often a salary drop)
  • is too comfortable without some additional income coming in

Too expensive? Maybe. But probably not: It removes a lot of the infrastructure of the benefits system (which is expensive) and results in more money coming into the system (more people able to find jobs, no tax-free band of income on those actually having jobs). I’m sure it would cost more than what we currently have, but I don’t think much more.

Worried about people “getting a free ride”? I don’t think they would: If anything I think this system would result in more people who are currently on benefits getting jobs. And even if they are, so what? Benefit scrounging is an exaggerated problem. It’s more important to encourage those who are genuinely willing to work than it is to punish those that aren’t – particularly when the cost of finding them is greater than the benefit of having done so.

I don’t know what to do about benefits for people with children. I need to think about it more thoroughly.

Edit: It’s been pointed out that this doesn’t cover disabilities for people unable to work due to disability. I don’t know the details of what I’d propose here, but I’m definitely for it

Healthcare

I cannot overstate how much I am in favour of socialized healthcare. Health is a right, not a privilege.

Making it a privilege is also really stupid. It results in you paying more per head (including taxes) and sacrificing herd immunity.

So, public healthcare for all.

I am also in favour of people being able to pay for better (or at least prompter) healthcare. Rich people get nicer things. Sorry guys, fact of life. If you don’t let this happen they’ll do it anyway on the sly.

What I don’t like about the current system in the UK is that the money spent on private healthcare does not (usually) benefit the public system. So here is what I propose:

For life-threatening or quality of life destroying issues, money doesn’t come into it. Everyone gets treated equally, no exceptions.

For other issues, there is a “banding” system: The resources are divided into multiple bands, and you can pay to move up a band. There is no difference in treatment between the bands, but the fact that you had to pay to be in the higher bands means that there’s fewer people in there so the service is less congested and the wait is shorter. The price points on the band are set dynamically based on capacity and demand.

(What follows next is the important bit)

All money spent on moving into a higher bands goes back into the system. The medical system is emphatically a not-for-profit. When you are paying for improved healthcare for yourself, your money goes directly to making healthcare for everyone better.

Medicine is very heavily regulated and required to be evidence based. “Alternative medicine” is as such generally illegal – it’s probably legal to sell homeopathic substances (because they’re indistinguishable from inert non-homeopathic substances), but making any claims of medical benefit will be met with severe censure if you can’t back it up with science

Education

I actually have quite a detailed plan for education. I won’t go into it too fully here, but here are some highlights:

  • No private schools
  • School attendance is free and mandatory.
  • All schools are boarding schools (however you get individual – albeit super tiny – rooms with en suite bathrooms. Dormitories for children are a thing of evil)
  • You don’t bring anything with you to boarding school. Clothes and equipment will be provided for you while there.
  • You do not choose which school you go to. It is randomly (qualification: For children with special needs of some sort it may be randomly assigned to you from amongst a smaller pool of appropriate schools) assigned to you. (THIS IS THE IMPORTANT BIT. I rank this point over most of the others)
  • It is massively illegal for individual schools to accept donations. Money may be donated to the school system overall (and for specific areas, scholarships and programs), and it will then be distributed fairly amongst the different schools.
  • Schools are not fixed term length – you make progress in different areas individually. Whenever you’ve amassed the criteria for graduation (which are “basic general competence in all areas, significant achievement in at least two”) you are free to graduate, whether you’re 12 or 21.
  • As well as current areas like languages, maths and sciences, other important areas include: Critical thinking (logic, rhetoric, research skills), physical disciplines (dancing, yoga, a martial art), life skills (cooking, cleaning, doing your own laundry). The idea is to produce well rounded human beings.
  • Students take part in a lot of the chores around the schools as part of the “life skills” section.
  • Programs of healthy eating and daily physical exercise are mandatory. What you do afterwards is your own business, but you will be in shape when you leave school.
  • There are no “classes” per se. Students work in small study groups of about 5 lead by a teacher. These are interactive and at least partly self-directed – the teacher is there to set the subject, keep things on track, keep things movement and to answer questions far more than they are there to lecture.
  • There aren’t fixed school holidays. They are structured much like work holidays are now – you have X amounts of time to take per year which you make book appropriately.
  • Students who are genuinely struggling get help. students who are simply unwilling to work, get flunked. They’re welcome to go out into the real world, find out how much fun it isn’t without a decent education and come back to school when they’re ready to take it seriously.
  • The teachers pay a lot of attention to social behaviour in the school. Study groups with unhealthy patterns get broken up, childrens’ rooms get moved around if needed, people behaving inappropriately get punished. (doing this well needs a much lower student : teacher ratio and good training. I’m ok with that).

Crime and punishment

Step 1 for reducing crime is, I hope, already covered: The benefits and education systems should help eliminating conditions of poverty and lack of education, reducing one of the biggest causes of crime.

Step 2 is decriminalization: I’m extremely pro decriminalizing “victimless crimes”. In particular I would be in favour of making both (most) drugs and prostitution legal. Even if you regard them as unpleasant or unethical (I regard them as unpleasant but ethical), the simple fact of the matter is they’re going to happen and it’s better if they happen above board, regulated and taxed.

So, hopefully we have a much smaller crime rate. What do we do with the crime rate we have.

First things first, as you might realise from the book I just reviewed, I’m very anti prison. I think the current prison system needs to be thoroughly abolished. Prisons may be a necessary evil for holding people pre trial and pre sentencing, but they’re not a valid sentence.

Things I would use for sentencing:

  • Corporal punishment. Yes, really.
  • Curtailed freedoms – curfews, tracking tags, reporting to parole officers, etc.
  • Community service
  • Fines (maybe. I need to think about this one)
  • Potentially, capital punishment for repeat offenders or crimes of sufficient severity – if the only way to deal with someone is to keep them away from the population permanently, I’m for the death sentence in preference to life imprisonment (which could be described as “slowly torturing someone to death over a period of 60 years”. Doesn’t sound like the humane option to me). I’m not thinking of this being the penalty for crimes of passion, I’m thinking more for e.g. serial killers and war criminals.

Anyone who can make a convincing argument that their crime was at least partly motivated by circumstance – lack of education, addiction, etc. and they are genuinely committed to changing that circumstance may be offered a reduced sentence for entering rehabilitation programs (adult education, drug rehab, etc).

Everything else

As you might have guessed from the random voting post, I have reasonably strong opinions about governance. However I don’t yet have a coherent picture of what to do about it so I’ve not included it here.

There are also probably other things I have opinions on that I’ve forgot to mention.

Questions?

This entry was posted in Uncategorized on by .

Judgement and Asymmetric Errors

I recently read (most of) Epistemology and the Psychology of Human Judgement.

It’s a good book. Well written and well reasoned. I stopped reading it mostly because I’m not in the target audience – it’s largely a polemic against what they call Standard Analytic Epistemology, which isn’t a viewpoint I can even imagine holding, let alone hold, so I’m not really needing convincing.

It does however have some interesting discussion of reasoning strategy. Unfortunately I think much of that discussion falls prey to a very common flaw that not enough people are aware of: Not all errors are created equal.

I shall elaborate.

One of the points the book makes (and which I’ve heard elsewhere) is regarding statistical prediction rules. The basic idea is this: In many cases, simple statistical prediction rules (SPR) have a lower error rate than human experts. This is often the case even when you give the experts the result of the SPR and allow them to selectively defect. Therefore in these cases you should always follow the advice of the SPR and not allow the human expert to intervene.

Seems sensible enough, right? How could you possibly argue that a system which has more errors is better?

Right?

Wrong!

The thing is, the error rate is actually an utterly useless number in most problems: What you’re actually interested in are the rates of specific types of errors.

Consider the case of triaging patients coming in with possible heart attacks. You do an initial triage, and everyone you think might be having a heart attack is passed on for treatment. Everyone who isn’t has to twiddle their thumbs for a while. This is a classic case where SPRs do better than humans.

What’s the interesting feature here? Well, you really want to make sure that you don’t leave anyone having a heart attack twiddling their thumbs. That would be bad. If you let a patient or two through who aren’t actually going to die any time soon, not that big a deal.

So the desirable solution is to treat everyone! Of course, that’s not so practical: You only have a limited amount of resources. You can’t deal with everyone. That’s why you’re triaging in the first place.

But suppose the result of the SPR is leaving you a bit under capacity – not a lot, but say you could handle another 20 or 30 patients without seriously impacting your ability to handle the current ones. What to do?

Simple. Let the expert pick those 20 or 30 people from amongst the people that the SPR told to twiddle.

This cannot decrease the rate of false negatives, no matter how bad the expert’s reasoning strategy is: That is, everyone who would previously have been seen under this rule is still seen. So no heart attack patients that would previously have got in will fail to get in under the new strategy. However some of those extra patients might actually be having a heart attack (assuming the SPR isn’t perfect). So this reasoning strategy is strictly better – it stays within the resource constraints and saves more lives – even if it has a higher error rate (it’s not obvious that it does, but given typical SPRs and typical experts I expect it usually will).

What we have here is a form of selective defection: You are given the results of the SPR and allowed to change them if you desire. The key feature is that we only allow selective defection in one direction. As long as that direction is the direction we care about most, and as long as this selective defection is bounded to not consume more resources than we have, it has to be an improvement.

Edit: A friend points out that SPRs are also used as a scoring mechanism: Rather than having them be binary yes/no you instead use them to rank the candidates and fill up to capacity. I think the point about asymmetric errors still holds, but certainly the above formalism doesn’t. I’ll have to think about it.

This entry was posted in Uncategorized and tagged , , on by .

Sleep as an Droid export data

This is possibly the single most niche piece of software I have ever written…

I use a sleep tracker called “Sleep as an Droid”. I wish to be able to get data out of it. Fortunately, it has a data export feature. Unfortunately, its data export format is a bit bizarre (it’s CSV, Jim, but not as we know it). So I decided to convert it to a tolerable-structured JSON output before using it, and have put it on the web for anyone to use. source is also available on github if you care.

This entry was posted in Code on by .