Unintended Consequences of OLPC's G1G1 Sales of XO Laptops

   
   
   
   
   

Wayan has been rightly concerned about grey market leakage of XOs from countries to the US, due to the high demand for One Laptop Per Child computers. Another ongoing concern has been in-country theft, misuse, and redirection of the laptops to users other than the intended children.

olpc icon
OLPC XO is unique for now

The first problem has been mitigated by the G1G1 sales in the US, and the theft problem is solved by both the bitfrost security platform and the distinctly green-colored and unique design of the laptop. Reason goes that if you see an OLPC-looking laptop being used by anyone who's not a child, it's been stolen or otherwise coerced from its rightful owner.

Unfortunately, the G1G1 solution plugging the developing world to first world leak has opened up a new leak. With this, it is possible for someone in a country where the children have XOs to legitimately buy a G1G1 laptop through a friend or organization in the US. In fact, many schools and non-governmental organizations worldwide may find US-based organizations to buy a small number of G1G1 XOs to use if their country is unwilling or unable to afford a mass XO purchase; it is unrealistic to think that the G1G1 laptops will stay only in the US.

This causes what turns out to be a non-insignificant problem. The social solution to theft was to eliminate the resale market value by distinctive branding - the only valid way to get a bright green OLPC XO was to be a student. Negroponte mentions this often in his speeches, and a 2005 Technology Review interview with him illustrates this "post-office truck" anti-theft method:

JP: How can you reasonably believe that these very valuable devices -- worth more [than] two working people's annual salaries in poor nations -- will not be stolen and resold?

NN: Having them stolen may become our distribution model, for all I know! Seriously. Usually people steal because there is a secondary market. There is not much of a secondary market for post-office trucks, so not too many are stolen. Also, imagine a UN blue rubber laptop, with the crest in it. How many of those will be stolen? I know, some will be, and people may even try to take them to a body shop to be transformed. (Technology Review, 2005)

Unfortunately, with valid ways to get an OLPC XO through G1G1, this loses its bite, as Mike C. Fletcher points out in this week's OLPC security email list:
Originally (to my knowledge), the plan was that some corporate partner would be contracting with Quanta to produce a custom run of the laptops, with some physical differentiation, such as a change of colour, so that the two "products" (the educational and the purchased) would be visibly different, possession of one type would be a badge of honour (those who help support), while the other would be a badge of shame (those who have supported the grey market and theft of children's laptops). That approach apparently was not feasible, so we wound up with a situation where our own program may be opening up a grey market.
The thread wanders into the classic gray market problems of children or their families selling their laptops for food, corruption in the laptop distribution chain, and outright theft with intent to resell that have haunted the XO since its inception.

Ivan Krstić's bitfrost does a laudable job at reducing these threats; given some important prerequisites of a national network (or well-maintained sneakernet with school mesh networks and a server) with a low level of corruption among the people managing it. The laptops must go through an extensive activation process during their first usage and then check in every two weeks (a value the country can alter), shutting off if they have been reported as stolen.

drunk coding
Another XO laptop shame

This is missing the point that Mike Fletcher is trying to make; if all the OLPC laptops look the same, regardless of if they were bought legitimately through the G1G1 program, received as part of a software project, or implemented by the country and given to the schoolchildren, then the social disincentive -- the badge of shame -- against theft is gone.

There are always social ways around the technical barriers that bitfrost puts into place, as Mike Fletcher reminds the thread:

Similarly if armed men are telling you to hand over the shipment of laptops and all of the activation keys, and tell you they will come back and kill you all if you ever report them stolen, you will likely hand over the shipment and keep quiet. As long as the profit motive exists, you will have people try to exploit the resource. [...]

What we are suggesting here is a means to reduce the *motive* to steal the laptops. While first-boot activation erects another hurdle (and we want that hurdle), we have potentially millions of dollars available for determined thieves. Having a physical difference in the laptop introduces a per-unit cost to grey-marketers, each laptop now has to be physically altered with a reasonable degree of care to avoid being easily spotted.

However, the physical alteration of the G1G1 laptops is not up for discussion. The list toyed with the idea of just a modified XO logo on the back, but it appears to be too late for even that; which would've been a difficult way to distinguish legitimate G1G1 laptops from stolen ones anyhow.

The more subtle the design differences are from G1G1 laptops and OLPC XOs as sold to countries, the less protection against theft there is. Using the design and coloration as a theft deterrent was thin ice to begin with; and requiring people to remember different color codes for the XO logo before branding someone as a thief more so.

So, a question to the OLPC News readership - is there a way to maintain the "mail truck" social immunity from theft by eliminating the secondary markets through distinctive branding of the OLPC XOs, and also allow identically-branded G1G1 systems to be sold?

It enables the argument of someone in a country implementing the XOs to claim that their XO was legitimately purchased; removing the stigma of having stolen it from a child. One early suggestion is a voluntary (?) registration list, available online, of G1G1 owners, but this doesn't help restore the "stigma" associated with an adult carrying an XO.

Related Entries

22 Comments

While I understand the argument for keeping two (or more) distinct physical packages for the laptops, this may also create a demand for the variants of the XO machines. Anyone who remembers the Swatch Watch craze of 1980's will get the idea.

Currently, people in the USA are willing to pay $400/XO. For $400, they are getting a significantly less powerful machine than a $500 special for a Dell laptop.

Collectors will certainly pay more for the XO laptop for the uniqueness of the hardware and the software inside. Others will buy because of the hubbub on all of major news sources. For whatever the motivation, the XO is going to have a demand in the 'developed' world.

This is the reason I'm giving for not doing the 'Give One - Get One' thing. I'd rather 'Give Two - Get None' and see two kids get a free laptop.

This is also the reason I'm buying myself an Asus Eee instead. No stigma there! - and far more practical (since I dont need a classroom full of other XO users to collaborate with).

Stigma? Seriously??? Maybe if I own one I finally start writing software for the platform. Maybe if someone sees it they will ask questions, etc...

Jon, while I find bitfrost quite an achievement, and would like to think that social stigma would be enough to prevent theft, I believe that's a rather simplistic understanding of crime in developing nations. It's not the case anywhere nor all the time, but there are many living in these places that are immune to stigma. I know of 12 year-old kids stealing sneakers from 10 year-olds in middle class neighborhoods in Lima as "deputies" of older people...

It's quite complex and certainly more than just a matter of differentiating computers. In a place where criminals steal police squad cars for the heck of it, and are ready to kill for very little money, well, bitfrost may not be enough.

Why should anyone be stigmatized having an XO laptop? This way the G1G1 program would be worth nothing (you are generous by buying a laptop to an African child, and then you are called a thief).
Even if you live in Africa. Only your inner circle would know that you are not a teacher or someone with legal right to have one, so nobody could cry thief only by seeing you with it.

This is an artificially created problem by inadequate distribution model used by OLPC. The XO should be sold everywhere for the lowest possible price and should look more like an ordinary laptop. Special and distinctive marking should be done by the local education program management, not in the factory (if those are bought in batches for an education program).

As the example of Austria proves it, the XO is as suitable for the developed world as it is for poor coutries. Actually, I believe it is much better suited, since it still has certain infrastructure dependence which is in several cases missing in poor countries (electricity for one: no personal electricity generator equipment for the XO yet, it was only promised).

Therefore, the OLPC program would be better off concentrating on how it could provide the most people with the XO instead of trying to give it only to children and putting up defences everywhere around to prevent happening otherwise.

Why should anyone be stigmatized having an XO laptop? This way the G1G1 program would be worth nothing (you are generous by buying a laptop to an African child, and then you are called a thief).
Even if you live in Africa. Only your inner circle would know that you are not a teacher or someone with legal right to have one, so nobody could cry thief only by seeing you with it.

This is an artificially created problem by inadequate distribution model used by OLPC. The XO should be sold everywhere for the lowest possible price and should look more like an ordinary laptop. Special and distinctive marking should be done by the local education program management, not in the factory (if those are bought in batches for an education program).

As the example of Austria proves it, the XO is as suitable for the developed world as it is for poor coutries. Actually, I believe it is much better suited, since it still has certain infrastructure dependence which is in several cases missing in poor countries (electricity for one: no personal electricity generator equipment for the XO yet, it was only promised).

Therefore, the OLPC program would be better off concentrating on how it could provide the most people with the XO instead of trying to give it only to children and putting up defences everywhere around to prevent happening otherwise.

I think this issue is addressed with the buy 100 and give 50 or the buy 1000 give 250 program recently announced.

While the distribution is scarce and demand is high, the laptop will be valuable and a grey market will emerge. The only long-run solution is to make the laptop as available as it can be. As demands meets offer, the laptop value will come down to it´s real price and no grey market will be necessary.

(as a note: being able to turn a $200 laptop into something a lot more valuable is a nod to the design success of the product)

Re: Stigma - it's not the US buyers with the stigma; sorry if I was unclear. What I meant to say that was for adults in countries with OLPC projects, it is now possible for them to own an OLPC legitimately, through the G1G1 program and a US buyer. Therefore, this adult did not have to steal an OLPC from a child to acquire it, especially if someone takes advantage of the "give 10,000" program and "donates" (re-sells) them cheaply.

Eduardo V-M: You are most certainly right; I've never been particularly convinced of the "mail truck" method of theft prevention myself, but I didn't want to get into arguing the details of bitfrost and branding and more than necessary.

Alexandre: the funny thing is that the branding approach to theft prevention will only work in the very short term. The children own the laptops, and if they can maintain them, will keep them once they've graduated and become full-fledged adults; so the "only students should have the cute green laptops" approach rapidly loses its effectiveness. Presumably, increased supply over the next few years after the launch will reduce the demand for theft (NB: I don't buy that).

"Alexandre: the funny thing is that the branding approach to theft prevention will only work in the very short term. The children own the laptops, and if they can maintain them, will keep them once they've graduated and become full-fledged adults; so the "only students should have the cute green laptops" approach rapidly loses its effectiveness."

Being able to spot stolen wares easily always helps to avert a black market. Not 100%, but all bits help. So in this light it is a pity that advantage is lost.

The "graduating student" effect can be solved by changing the designs of new roll-outs. So the cute green one is the 2000 noughties version, the orange one is the 2010 rollout, etc.

Winter

What makes this laptop unique to theft? Aren't all laptops stolen if one is not careful? It's not drugs we're talking about. Is there any other product that is high jacked or stolen from kids? The mail truck thing is stupid. Its more like a mac thing, they all look exactly the same and are registered online so you can report them stolen. Does it do any good i don't know but i do know you cant stop crime so who cares if these computers get stolen? everything does. Its better to have the laptops all over the world than not.
Word.

Zak wrote:

"Its more like a mac thing, they all look exactly the same and are registered online so you can report them stolen."

Gamoe writes:

Huh? Registering your Mac is voluntary, and like any other product registration, which most companies do. I don't see how it's a theft-prevention method, nor how it makes the Mac unique in this area in any way.

Anyway, it is a little bit of a hole in the "mail truck" idea, and I wouldn't have minded an XO with a differing design for the G1G1 offer.

Robert Arrowsmith opined:
"I'd rather 'Give Two - Get None' and see two kids get a free laptop."

Go to http://www.laptopgiving.org/ , click "Give a laptop now." on the right which leads to a PayPal donate button. One might assume that you're talking rhetorically out of your ass and have no intention of making a positive contribution (like everyone else who pisses on the OLPC effort), but I try not to be cynical about people just as I'm optimistic about the OLPC.

Sorry for being late to this conversation. My intention is to G1G1 and use the laptop I get for myself. I have told a lot of friends about it, focusing on people who have kids or grandkids but I like the machine better than the alternatives available in the US. I am 62 and have no kids of my own. Am I doing wrong ? As far as I can see, the XO would be a good laptop for my needs and at $400 cheaper than the equivalent I could get with more interesting features. Am I doing something wrong ?

Marc,

The problem isn't at all with consumers using the G1G1 program; buy away -- I think it's a great option for you. The problem lies with the market dynamics that OLPC has enabled through the G1G1 program, reducing the efficacy of their theft-prevention plans. Ideally, they would have branded the G1G1 laptops that you'd receive dramatically differently from the ones shipped to countries as part of educational programs.

I'm sure other people have tossed this idea around, but why didn't the OLPC decide on biometrics as another security feature. Was it simply too expensive to include? USB Fingerprint Reader hardware can be found for $10. Though this system level feature will not prevent people from scrapping the parts like the dual mode screen unless a biometric key mechanism could render the screen useless. All features that would probably cost too much...

Or is the source of the theft at the distributor level? I would think that with the introduction of Brightstar that theft would be eliminated effectively at that level.

i can see how it would be difficult to hold on to an expensive looking piece of equipment if your family has barely enough to eat.

on the other side, i'd like to get into the G1G1 program, without being branded a thief.

Excuse me for being rude, but i'd like to keep the device for myself (or my kids) - it would be one heck of a ebook-reader with higher resolution & DPI specs than current dedicated eBooks (including Sony's or Cybooks gen3 reader) - especially with the screen rotated to face top-up.

I have no problem using the XO laptop in public, it's just that i don't like to be branded thief

i can see how it would be difficult to hold on to an expensive looking piece of equipment if your family has barely enough to eat.

on the other side, i'd like to get into the G1G1 program, without being branded a thief.

Excuse me for being rude, but i'd like to keep the device for myself (or my kids) - it would be one heck of a ebook-reader with higher resolution & DPI specs than current dedicated eBooks (including Sony's or Cybooks gen3 reader) - especially with the screen rotated to face top-up.

I have no problem using the XO laptop in public, it's just that i don't like to be branded thief

First, re hijacking. You send the laptops and the activation keys separately. Truck drivers can't hand over information they don't have.

The real problem is war zones, where the schools themselves can be threatened. I don't know whether Shining Path activity rises to that level anymore. Anyway, as far as I know, cocaine is far more profitable than laptops.

In war zones the problem is probably not going to be theft but destruction and targeting of children. The Taleban, for example, have been destroying government schools in some southern parts of Afghanistan. Rebels aren't going to allow thousands of children with video cameras to upload their mayhem to YouTube.

PS You can get a Live CD of the XO software if you want to give your kids more powerful computers. It runs on any x86 PC, including recent Macintoshes. It's on the Live_Cd page on the Wiki. The current ISO is somewhat out of date, but there should be a new one after the chaos of the production ramp-up subsides. The Live CD includes a Gnome desktop, but doesn't have the mesh networking capability. It's certainly worth a look.

Edward, I'm pretty sure Peruvian officials are not going to send the computers into troubled areas. In some regions of Peru, both Shining Path and drug traffickers are attacking openly police detachments and barracks; it would be mad to send anything there. And there's no connectivity in those places. Everything points towards implementation at Andean schools, with other kind of security problems but with connectivity set up by old projects and better conditions (still brutally poor, though).

Why don't they just make a way to legally buy an XO instead of making people outside of Canada and America jump through all kinds of hoops to get one? One don't they just extend the time limit so you can buy one legally next spring when you finally see a real one for the first time?

Giving to caritative organizations is always a problem. Part of the money involved is always used to pay someone doing the job along the path of delivery, some is lost, some is stolen - how much is getting to *help* someone is not clearly known.
There is always someone deciding where the moneys will go to, specifically - not a personal choice of the giver.
Still, those that have chosen to give can make choices: health (research, facilities), education (books, equipment, expertise), catastrophe relief, etc. It can be local, national, international.
The question remains: do we keep looking for the *perfect* situation or do we give some of our own, now, hoping for the best?

I say we in the "first" world strive to flood the rural schools in the "third " world with laptops. With the consequence that everyone who wants one can get one. Some places will have no theft, others will be really terrible so they don't have to responded to equally.

Just an idea.
I really like whats going on here, it helps affirm my faith in Humanity.

Close