SecondLife? is a system where people use clients (computer applications) to interact with (and via) a virtual 3D world (a "sim") hosted on remote multi-user servers. In this world the person is represented by a 3D object (an "avatar") whose movements within the 3D world the person directs. Not only can people move their avatars around and talk with other avatars on the sim, they can also edit the sim to create new 3D objects, and write scripts to control how these objects behave. These objects can be traded between players for money (there's an in-game currency "lindens", but it is freely convertible to real world dollars or pounds).
One way of looking at SecondLife? is as a preview of how economics might work in the real world, if everyone had access to a 3D printer, and the price of raw materials reduced sufficiently to no longer be a bottleneck.
While everyone in SecondLife? can create their own clothes, swords, body shapes, etc. most people don't. It takes skill to make clothing that looks good and hangs naturally, or to write a script to make sword fighting interesting. Most people, rather than settle for inferior looking clothes they make themselves for free, are willing to pay money to others who have specialised and invested the time to hone their skills at producing that particular type of object. They get the lindens to pay for these nice clothes either by trading their own swords or services (such as sitting around in someone else's sim to increase how popular that sim appears to be), or by buying lindens with external currency such as dollars that they earned at a 9 to 5 job selling burgers in the real world.
Dyson predicts the impact of 3D printers is not that everyone will print out a new power cable for their Apple iPhone when they lose one, but that everyone could do so which, through supply and demand, will drive down the price and stop Apple charging ten times the cost to them of producing the cables. She also predicts that, in practice, there's a break even point somewhere between transport costs and the advantages of bulk manufacture, and the shorter the production run or simpler the object, the closer to the home the optimal point for that item will be, so '3D print shops' will spring up in villages or towns, where, like in SecondLife?, the cost of the object will depend more upon the attention required from an expert creator to produce it, than upon the cost of the raw materials.
This ties in with something she talks about in her essay on the AttentionEconomy?. There is a trend in history to increase commoditisation - to increase the number of different types of thing that can be traded in a market to receive a price that more accurately reflects the value of how in demand that thing actually is. Standardisation (a standard carton of orange juice or a standard size of gold bar) is one means of increasing reliability - the ease for a distant purchaser to decide to risk paying for something without having seen or experienced that particular object or service before. The easier it is for purchasers to rely, the lower their overheads (they don't have to hire an assessor they trust to physically travel to you to look at the object on their behalf), so demand and the size of the market increases, and so does the value the provider can get.
A man who makes flint axes for hunters in his own village will get less in return than a man who is able to sell those same axes world wide on eBay. A man who cares for his elderly mother will get less tangible returns than one who is a qualified care assistant whose services are being bid upon by two competing care home companies in his area. A woman who works as a £50 an hour lawyer and uses some of her salary to pay for a daycare provider to look after her children (which uses £10 an hour assistants to look after 4 children per assistant) during the day, has an economic incentive to do so. (Indeed, people can debate whether the move into employment of young mothers benefits society, but there's little debate that the increased average income of couples has been so significant that it has even driven house prices up).
In her essay, Dyson talks about the impact of the internet not just being that it changes the balance of power between individuals and institutions, but that it is allowing people to commoditise their attention - to explicitly sell their eyeballs, rather than just let third parties hoover up that value by packaging and reselling attention to advertisers.
She writes:
Will the world of cognitive surplus make it easier for us to be environmentally responsible, guided also by the inclusion of externalities in the prices of physical goods, so that we end up “consuming” fewer physical things and spend more on virtual value?
This attention economy is not the intention economy beloved of vendors, who grab consumers’ attention in order to sell them something. Rather, attention here has its own intrinsic, non-monetizable value. The attention economy is one in which people spend their personal time attracting others’ attention, whether by designing creative avatars, posting pithy comments, or accumulating “likes” for their cat photos.
The "non-monetizable" is an interesting qualifier, there. Perhaps it would be more accurate to say that it is not always efficient to transfer between currencies. You can, for example, find ways to convert EveOnline? in-game currency into real world money, but this is (deliberately) difficult to do and so, although it is "monetizable", turning it into money has sufficiently high overheads that, in practice, the two economies and functionally seperate. Their economic cycles are not in lock-step, and those who are rich in in-game currency are not particularly likely to be richer in real life than a fellow player who is poor in-game.
In her third essay, Dyson talks about networking. Some people go to conferences with a 'hit list' in mind of people they want to network with - talk to, exchange business cards with, impress, and eventually sell their product or idea to. She points out that for most people it is unrealistic to go to the forum in Davos and aim to spend all your time there with Bill Gates, Nelson Mandela and President Obama, because you are probably not the person they want to spend 1/3 of their time listening to. Their time and attention is more valuable, in the 'open market' than yours is.
She suggest that the effective thing to do is to work your contacts and find people you already know (and who perhaps owe you favours, or at least like you and are already impressed by you) who can effect an introduction. Or, to put it another way, if your aim is to 'buy' 10 minutes of attention from target X, and X's attention on the open market is 6 times as valuable as your attention is, then it is worth taking 60 minutes of your attention and 'spending' it by making presentations to and gaining endorsements from people who person X is already impressed by or owes favours to. And, perhaps, a further 60 minutes in listening to and endorsing the ideas of those people, to keep the accounting even.
A recommendation that can be purchased is worthless, like a book jacket endorsement from an author who is a nice chap and will provide a quote "Comparable to Tolkien" for the book of anyone who buys him a pint, no matter how terrible the book.
So what's really of value here is time and attention from a person who not only has the competence to judge an idea or project, but who also has the integrity to only endorse the things they think actually merit being endorsed. This goes beyond the web of trust I talked about in /MediatedTrust, because there's an economic aspect to endorsement. The endorser is saying not only that they have reviewed piece of music "foo" by musician "bar", and think that in the music genre of "bat" it rates 7 stars out of 10. They are selling to musician "bar" 30 minutes of their time in return for writing a review, with no guarantee that review will be positive, and the value in the review is that (if it turns out that it is positive) that endorser is not just providing a rating, they are staking some of their own reputation. They are saying to a particular audience who trust that endorser "I think you will find this piece of music is worth spending 30 minutes of your time listening to". Or, in a different context, "I think this kickstarter project is worth your buying into, and that it is going to take off and be a success."
Currently networking is a mainly non-commoditised activity. You may be able to buy a chance to sit at a table with a particular group of people, by paying a high price to go to a conference they are at, where the conference organisers know they are able to charge that price in part because of the networking opportunities being afforded by the event. But this can change. The relentless trend of enabling people to squeeze more value out of their time, by commoditizing things and selling them in a market, will sooner or later take advantage of technological systems that allow people to automate saying "I have people's ears in speciality Z1, and I've got 3 hours free this evening. If you've got something you think I would consider to be worth endorsing to people interested in Z1, how much of my time do you want and what are you willing to offer me for that time (no guarantee I'll actually endorse, though)?" and get matched up with someone who has a path to the ears of speciality Z2 that the first person in interested in reaching, in exchange.
Imagine going to a standard conference, but wearing GoogleGlasses?, and using an app that does facial recognition, and can calculate paths of interest, matching you up with other users of the same app to arrange mutually beneficial meetings.
(Note: Edward Snowden recently (Mar 2014) attended the TED conference as a robot. Given an ability to multi-task, the way people spread their attention on several chat channels at once, there's no reason he couldn't have attended as two telepresence robots. Imagine an expert knowledge system, controlling a [vactor], that handles basic interactions on someone's behalf but knows when to grab the real person's attention to handle someone beyond the system's ability to cope, or that would be of particular interest to the person.)
[6 degrees of separation] suggests that an acquaintance chain of only a few links can be found between most people. There's a use for that in cryptography.
The problem with secure end-to-end encryption is doing an initial key exchange in a way that's secure even if there's a man in the middle. For example, let's consider Liao Yiwu, a poet living in China who the Chinese government have banned from travel, because he wrote a poem about Tiananmen Square. Suppose Liao wants to publish a new poem in the New York times. From a paper copy of the times, he discovered the email address of the editor of the Time's 'On Poetry' column, David Orr <dorr@times.com>. Liao decides to email his poem to David, and ask David to publish it anonymously, so Liao doesn't get arrested. Liao doesn't know anything about David that isn't public knowledge, and has never met him in person.
Suppose Liao uses his laptop to generate a public-private key pair. If he can get ahold of David's correct public key (the one that matches dorr@times.com) he can use that to encrypt his poem with it, and send it to David without risking leaving proof with the Chinese authorities that he, Liao, is the author of the poem that gets published, because only David will have the private key that's able to decrypt a message encrypted using David's public key. And if, along with the poem, Liao sends David a copy of Liao's new public key, then David can use that to securely reply, thereby setting up secure two way communication.
But how can Liao obtain the correct key? If he sends a request to a public-key server saying "Send me the public key for dorr@times.com", the Chinese government might intercept that message, and substitute a key of their own in the return message, acting as a man in the middle. Similarly, if he looks for certificates attached to David's key that say "David attended a key signing party with me, and I attest that this is his correct key", those too could be faked by the man in the middle, unless Liao has a way to verify that at least one of the authenticating keys is genuine.
But what if Liao hasn't? What if he has met a few other dissidents in China in person, but doesn't have their keys, nor the keys of anybody else, other than his own key?
One solution for this would be to Liao to do a brief Skype call to someone who:
(A) Liao does know well enough that an actor couldn't replace him; and
(B) Who does have reliable knowledge of a public key that Liao could use to bootstrap
The person phoned could read the public key out aloud to Liao. Or, if Liao had been able to get hold of the key, but just wasn't sure it was the correct key rather than a fake substituted in by the man in the middle, Liao could hold up a sign with the fingerprint of the key, and the person Liao called could look at the fingerprint, compare it to one on their own screen, then tell Liao either "yes" or "no".
And here's where /MediatedEndorsement and 6 degrees of freedom comes in. If Liao doesn't know such person, he needs a way to work through the acquaintances he does know, exchanging keys at each step, until he reaches someone his contacts trust who is on the main non-Chinese key signing network. And he needs a way to trade favours along the same path, so that random stranger is willing to spend a minute or two looking at keys over a skype call. If both people on the call have the same software installed, possibly the [Socialist millionaire] could be used.
Sina Weibo already has much of the relationship information needed.
In theory, if the WebOfTrust are signing documents at a sufficient rate, that the security establishment of your country couldn't keep up, in real time, without lag, then you could wait for a pre-announced release time (eg a daily build), and try to download it immediately, then if there wasn't time for the man in the middle to decrypt the whole thing then re-encrypt it with a false key (because you started to download it before the end bit had been released), such as a nightly build, you could trace that key back to the key you wanted.
![[Home]](/img/text_logo.jpg){kind=logo}
DouglasReay/MediatedEndorsement