?

Log in

No account? Create an account
Stuff - Jackdaws love my big sphinx of quartz [entries|archive|friends|userinfo]
Scott

[ userinfo | livejournal userinfo ]
[ archive | journal archive ]

Stuff [May. 8th, 2010|01:55 am]
Scott
[Tags|]

Here are two of my Ideas For Society Optimizing Websites That I Would Make A Fortune Off Of If I Were Some High-Powered Go-Getter Silicon Valley Entrepreneur Person, Which I Am Definitely Not, Even Though I Got An Article Of Mine Near The Top Of Hacker News Once And Will Probably Never Let Anyone Forget It:

Start a site (or app) to organize boycotts and petitions. Boycotting is a coordination problem. If a million people boycott (to take a topical example) BP, then BP will be forced to change its evil ways and everyone wins. But on an individual level, boycotting is always a net loss - it's inconvenient for me to drive to a different gas station, but the loss of a single person's business one way or the other won't affect BP at all. This is why a company can do horrible things and get away with it - and one of the reasons why I disagree with libertarians who say the market will naturally take care of evil companies.

What we need is a site where I can promise to join a boycott if and only if the boycott gets a certain number of members. For example, I register and "friend" the BP boycott page, which has been pre-made by some activist to read "We want 1 million people to boycott BP until they donate a billion dollars to cleaning up oil spills." I don't immediately have to do anything. But once the millionth person signs up, the site sends me an email saying "Your BP boycott is now active; please stop buying BP products." All of a sudden, BP has a huge problem and will probably give in, at which point the site can send me an email saying I can stop boycotting them.

This is useful to consumers, because it removes the risk from boycotts - they don't have to join it unless it would work. It's useful to organizers, because they can convince and organize consumers more easily. And it's useful to corporations, because they know exactly how many people are boycotting them at a certain time and why - and because if they're smart they'll do whatever it takes to prevent that page from getting a million people!

Along with boycotts, the same mechanism would help get donations to charities and political campaigns (ie "I'd like to donate to the Libertarian Party, but they have no chance of winning if only one person donates, so I'll donate $100 if and only if I know they can raise $10 million." And then once the number of pledges gets above $10 million, it'll activate all those pledges, and all the Libertarians can donate knowing that their donation is going to be useful and their party actually has a chance of winning.

(the same could very well be done with votes, but I don't know if it would be legal. The government tends to get kind of touchy when voting is involved.)

Again, there's the problem of enforcement. The honor system would probably work pretty well here, but you could get a deal with PayPal to give an option to verify anything that involves paying money.

A somewhat primitive example of this kind of thing are Facebook groups like Once we reach 4,096 members, everyone will donate $256 to SingInst.org. I use the plural "groups like", but this is the only one I've actually seen (I haven't really been looking).

And speaking of Facebook, my second idea is a Facebook dating app. There used to be a site where someone, let's call him Bob, would put in the email address of a crush, let's call her Mary. Then it would sent Mary an email saying "Someone likes you! Enter the email addresses of the people you like to see if there's a match." Then if Mary entered Bob's email, it would send an email telling Mary and Bob that they both liked each other, at which point they could have wild passionate sex.

There were a few problems with this site, most notably that it was a STUPID ROTTEN SCAM. But the concept was sound. Whether or not to tell a crush you like them has a payoff matrix that discourages people from trying, which leaves everyone worse off. This method would have eliminated the negative payoff if it hadn't all been a STUPID ROTTEN SCAM.

With Facebook, we have an opportunity to finally realize the full potential of this technology. Imagine an app where you can secretly mark Facebook friends with a "have a crush on this person" tag. If the app detected two people who had both tagged each other, it would send them a notification. Short and simple. A more heavily customizable version would allow settings like "Not going to initiate anything, but if this person is really into me I'd be willing to try a few dates" versus "Really, really into this person" - such that if both people had the "not initiate anything" setting it wouldn't trigger a note, but if one person had the "not initiate" setting and the other had the "really into" setting it would.

The only potential problem is idiots who tag everyone on Facebook so that they can see who tagged them and make fun of people with unrequited crushes. One could probably solve this by having the app auto-announce any matches on the feeds of both parties' friends; hopefully the potential embarrassment factor would keep people honest. This is, of course, only a short-term solution; the long-term solution is taking everyone who would deliberately ruin a useful app so they could make fun of people who are honestly interested in them, and throwing them into a pit of poisonous snakes.

If anyone reading this is a High-Powered Go-Getter Silicon Valley Entrepreneur, I grant free use of these ideas to whoever who wants them.
linkReply

Comments:
[User Picture]From: alicorn24
2010-05-08 01:17 am (UTC)
http://fundable.com/ The money version exists, or at least used to.
(Reply) (Thread)
[User Picture]From: squid314
2010-05-08 01:59 am (UTC)
"Confound these ancestors! They've stolen our best ideas!" - Ben Johnson

I looked for the boycott site to see if it already existed, but didn't think to look for the fund version. Disappointing to see it out of business.
(Reply) (Parent) (Thread)
[User Picture]From: monolith94
2010-05-08 03:21 am (UTC)
They're already doing that first idea you have with coupons.
(Reply) (Thread)
[User Picture]From: xuenay
2010-05-08 04:00 am (UTC)
The boycott site idea is awesome.
(Reply) (Thread)
[User Picture]From: cakoluchiam
2010-05-08 05:08 am (UTC)
I swear to god I saw this website a couple years ago and I thought it was called the Tipping Point, but nothing on Google seems to link to anything of the sort.

It was designed for money gathering (much like fundable) but with a slightly different system that iirc also allowed for abstract promises like "I will wear a green hat on tuesday if 7,000 other people also will".

If it actually did exist as I remember it and it did get vanished from the internet as appears to have happened, my best guess is because of the potential for its contracts to infringe on freedoms like voting etc.

It looks like fundable went the same way (or maybe fundable -is- the website I'm thinking of and its UI is just so drastically different from what I remember that I don't recognize it), so I'm sure there's some reason why social action sites such as those aren't more prevalent.
(Reply) (Thread)
[User Picture]From: cakoluchiam
2010-05-08 05:16 am (UTC)
Also, with regard to the facebook crush thing, I highly suggest you check out OKCupid's quickmatch system. It seems to work very close to the way you describe, though without such fine-tunable customization. Though the system itself presents as an app which shows you random users one at a time to rate, I believe it is also tied into the star rating system that exists on every OKCupid profile, so if you can find the OKCupid profile of someone you have a crush on, you can rate them there and quickmatch will do the magic for you.

The main downside is that people who are on OKCupid have already taken that first step toward admitting desire for romantic connection, which is probably a bit too much for the demographic which would want a facebook app of similar function.
(Reply) (Thread)
[User Picture]From: squid314
2010-05-08 11:21 am (UTC)
I've been using OKCupid for years, and I had no idea there was any sort of secrecy involved to the rating system. I need to check this out further.
(Reply) (Parent) (Thread)
[User Picture]From: ciphergoth
2010-05-08 07:14 am (UTC)
Is this proposal different from http://www.pledgebank.com/ ? It was Pledgebank that resulted in the Open Rights Group being founded...
(Reply) (Thread)
[User Picture]From: squid314
2010-05-08 11:20 am (UTC)
No! That's perfect! It's sort of disappointing how limited the pledges are, but hopefully there's room to grow. I'm totally signing up for this.
(Reply) (Parent) (Thread)
[User Picture]From: minion_for_hire
2010-05-08 11:08 am (UTC)
Suspicion that it would be a dirty rotten scam would keep me from using a facebook matchmaking service, but I do see the appeal.

On the other hand, the idea of voting for someone and only doing so if the votes will count is pretty much how Proportional Representation/Single Transferable Vote works. It's something that has been highlighted by the recent UK elections: 6.8 million people voted for the Lib Dems, and they got 57 seats. The Labour party got 8.6 million people to vote for them, and won 258 seats. Surely there is something wrong with that.
(Reply) (Thread)
[User Picture]From: squid314
2010-05-08 11:20 am (UTC)
Agreed, but it's probably easier to make a website than to change the voting system, at least in the US.
(Reply) (Parent) (Thread)
[User Picture]From: minion_for_hire
2010-05-08 11:27 am (UTC)
This is true. The one the bugs me is people saying that you'd be throwing your vote away over here, with our STV system. There is no such thing!
(Reply) (Parent) (Thread)
[User Picture]From: cakoluchiam
2010-05-08 04:54 pm (UTC)
In college I attended a very interesting lecture on the mathematics of voting systems where the speaker demonstrated how just about every system of voting can be broken or abused (e.g. where someone can win an election by getting the fewest number of votes). It was very interesting, but it also shattered any hope I had that a change of voting system would fix any problems with democracy.
(Reply) (Parent) (Thread)
[User Picture]From: sniffnoy
2010-05-08 05:22 pm (UTC)
That all systems have vulnerabilities doesn't mean that other systems can't fix existing problems (vote-splitting in particular). Also, if you're willing to consider cardinal voting systems, range voting basically has... well, no problems, really, IINM, aside from being cardinal.
(Reply) (Parent) (Thread)
[User Picture]From: cakoluchiam
2010-05-08 05:42 pm (UTC)
Nope, that has plenty of problems.

First off, most people are too lazy to consider the difference between a score of 80% and a score of 70%, or, too forgetful to remember whether they rated the candidate on page 3 that they liked a little better than this candidate on page 8 70% or 75% and thus whether it was safe to rate this person 70% or should they dip him down to 60%? Range voting requires every voter to intelligently consider not only the rankings of their choices but the weight of the differences in rankings, which you can't just jump into a ballot box and write down on the fly; you have to either consider it strongly before hand or you have to have an interactive voting aide available, which completely disenfranchises those who require absentee ballots because they can't make it to their local polling station.

Next off, there's the issue that for whatever reason, some people will just vote candidate X at 100% and all other candidates at 0%, which returns us to all the issues of current multiple-choice voting, but also creates the situation where the weight of these sorts of votes which are likely to come out of the less informed voters will count less than the weight of the votes from informed and thoughtful voters who actually place most candidates somewhere between 20% and 80% on their rankings, so the election will end up being swayed by the stupidest voters.

If we limit ourselves to strictly ordered voting (e.g. first choice, second choice, etc.), then the outcome of the election is entirely determined by the weights applied by the people who are designing the method of counting the ballots. Instant-run-off counting gives situations where the candidate with the fewest positive votes can win. Weighted counting gives different outcomes with different weight curves...

No known voting system is without flaws (and if you disagree, I might remind you that a voting system which is perfect if executed properly but is complicated for the layman to properly execute has the flaw of being unattractive and is likely to be considered too much of a hassle for everyone to take seriously)
(Reply) (Parent) (Thread)
[User Picture]From: sniffnoy
2010-05-08 11:40 pm (UTC)
If we limit ourselves to strictly ordered voting (e.g. first choice, second choice, etc.), then the outcome of the election is entirely determined by the weights applied by the people who are designing the method of counting the ballots. Instant-run-off counting gives situations where the candidate with the fewest positive votes can win. Weighted counting gives different outcomes with different weight curves...

Well, if we limit ourselves to ordinal voting, Arrow's Theorem applies! But that doesn't mean every ordinal voting system is as pathological as IRV (which isn't even monotonic) or variants of the Borda count (which has the teaming problem). There are some pretty good ones out there... I remember reading about Ranked Pairs and its variants a while back (though I don't think they call it that anymore?)... don't know what people have come up since then, but they get a lot better than the ones you are suggesting. Though can't get away from Arrow's Theorem, of course.

Next off, there's the issue that for whatever reason, some people will just vote candidate X at 100% and all other candidates at 0%, which returns us to all the issues of current multiple-choice voting, but also creates the situation where the weight of these sorts of votes which are likely to come out of the less informed voters will count less than the weight of the votes from informed and thoughtful voters who actually place most candidates somewhere between 20% and 80% on their rankings, so the election will end up being swayed by the stupidest voters.

Firstly: Is approval voting so terrible? I'm not seeing why it's a huge problem. Even if you don't particularly like approval voting, it's a decided improvement over plurality. (Also don't forget, range voting allows you to just not rate someone.) Secondly: That's not just "for some reason" - rating everyone 100% or 0% is (IIRC) the strategic thing to do. So, may depend on what you mean by "better informed". Of course, we're assuming that a large chunk of the populace will vote honestly - because in tests, they have! Looking it up, general result seems to be that probably ≤33% of people will vote strategically? (Note that under range voting, voting strategically does *not* result in putting candidates in the wrong order, just collapsing order/ratings.)

No known voting system is without flaws (and if you disagree, I might remind you that a voting system which is perfect if executed properly but is complicated for the layman to properly execute has the flaw of being unattractive and is likely to be considered too much of a hassle for everyone to take seriously)

IIRC, implementations of range voting for current voting machines already exist. It's also something that people are already familiar with from other contexts, and it doesn't require that you rate every candidate. I think you're overstating how complicated it is.

Range voting requires every voter to intelligently consider not only the rankings of their choices but the weight of the differences in rankings, which you can't just jump into a ballot box and write down on the fly; you have to either consider it strongly before hand or you have to have an interactive voting aide available, which completely disenfranchises those who require absentee ballots because they can't make it to their local polling station.

I don't see how your comment about absentee ballots is true at all. Voting absentee means you can take as much time as you want - look up the candidates on the internet while you're filling out your very ballot. I do that, certainly. It's an advantage to absentee voters, not a disadvantage. Or you mean having the interactive voting does, because absentee voters won't have access to it? Well, just put it up on the internet then!

In any case, do most places not send out sample ballots to people beforehand? Perhaps not; I wouldn't know. Well, as long as they do, I see no reason a person voting on location shouldn't have their votes prepared beforehand.
(Reply) (Parent) (Thread)
[User Picture]From: michaelkeenan
2010-05-08 06:41 pm (UTC)
I'm guessing that speaker was discussing Kenneth Arrow's Impossibility Theorem?

http://en.wikipedia.org/wiki/Arrow_theorem
(Reply) (Parent) (Thread)
[User Picture]From: minion_for_hire
2010-05-09 12:24 am (UTC)
You're arguing against a strawman there: To say no system is without flaws is prefectly reasonable. This is not to say that it is impossible for one system to be an improvement on another. Similarly, some systems are harder to break than others.

To take a bit of a specific situation, I offer an example like this one. http://news.bbc.co.uk/2/shared/election2010/results/constituency/b46.stm The Conservatives got 38% of the vote, Labour 30% and the Liberal Democrats 27%. Clearly, the voters of this constituency lean to the left (57-38%) but as the left-wing vote is split between two parties, neither gets in.

Single Transferrable Vote is relatively simple to explain (number the candidates in order of your choice) and generally works out in close enough proportion to the voters desires. http://en.wikipedia.org/wiki/Irish_general_election,_2007#Result shows the correlation between percentage of seats and percentage of first preference votes in the last Irish General Election. By comparison http://en.wikipedia.org/wiki/United_Kingdom_general_election,_2010#Results shows a complete disparity between the percentage of votes and the percentage of seats, most dramatically for the Liberal Democrats.

To my mind, this makes it a better system than First Past the Post as it makes it more likely to reflect the desires of the people.

By the way: I'm not arguing that it's a perfect system. I'm just saying it's better. http://en.wikipedia.org/wiki/Dublin_Central_%28D%C3%A1il_%C3%89ireann_constituency%29#2007_general_election is a specific example of it going a bit skewy, when Cyprian Brady was elected with a record minimum of first preference votes.
(Reply) (Parent) (Thread)
[User Picture]From: michaelkeenan
2010-05-08 04:52 pm (UTC)
You have reinvented the assurance contract!

http://en.wikipedia.org/wiki/Assurance_contract

The Wikipedia article links to five websites that try to implement them, including Pledgebank which another commenter mentioned (I think that's the most famous one).

You might be interested to learn about two other assurance contracts on Facebook:

"Once we reach 1 million members, every one of us will donate $100 to defeat aging!"
http://www.causes.com/causes/295774

"Once we reach 10,000 members, every one of us will donate $100 to defeat aging!"
http://www.causes.com/causes/297361

That second one has 1833 members. I check it from time to time and if the trend continues I think it'll have 10,000 in about 8-12 years. That's longer than I'd like, but on the other hand, it will happen, and it will be a historic event - the world's biggest assurance contract ever. Hopefully it will be considered a newsworthy event. I would love it if more public goods were funded in this way.
(Reply) (Thread)
[User Picture]From: cakoluchiam
2010-05-08 04:58 pm (UTC)
AHA!

http://www.thepoint.com/

That's the one that I saw! Thanks you wikipedia!

The site uses the phrase "The Tipping Point" as part of their marketing, but there is no "tipping" in the name.

Edit: unfortunately their browse feature appears to be borked for me...

Edited at 2010-05-08 05:01 pm (UTC)
(Reply) (Parent) (Thread)
[User Picture]From: squid314
2010-05-08 10:11 pm (UTC)
I am both delighted to know this has already been invented, and irritated to know that this has already been invented and people generally aren't using it.
(Reply) (Parent) (Thread)
[User Picture]From: michaelkeenan
2010-05-08 05:01 pm (UTC)
You'll be happy to learn that your article wasn't just *near* the top of Hacker News - it was in first place when I checked Hacker News yesterday.

There's a twitter feed of all the articles that hit the top of HN here: http://twitter.com/tophackernews
(Reply) (Thread)
[User Picture]From: squid314
2010-05-08 10:06 pm (UTC)
Wait, what? Something of mine was on Hacker News yesterday? I didn't know there'd been anything on there since that article on excuses.
(Reply) (Parent) (Thread)
[User Picture]From: michaelkeenan
2010-05-09 01:50 am (UTC)
Arundelo posted your blog post about people's lives being more messed up than you'd think:

http://news.ycombinator.com/item?id=1329296

It reached number one!
(Reply) (Parent) (Thread)