I've been sitting here for 10 minutes or so trying to find the right way to critique not just the article, but also the tone of the discussions here. I think the best thing that I can say is that this experiment flies in the face of everything that pg, Steve Blank and the other leaders of Lean Startup methodology are teaching.<p>If the customer here is a member of the "urban poor" (or what some commenters have decided to call "bums"), then I'd suggest that we start with some customer research. Do the jobless urban poor want to stand at a terminal in the street and perform menial tasks for spare change? Would this solve real problems for them on a daily basis? What happens when it's cold and snowy out? How many of them know how to use a computer, or read?<p>Buying into this concept as a way to solve urban poverty is no different than some Marketing VP sitting in her 50th floor corner office, thinking that she knows what her customers want from her company's web site without ever asking them. Before you start throwing out solutions to urban poverty, you might want to at least talk to one or two of the people whose problems you are trying to solve.
OMG! You guys! Aside from a few, most comments here are discussing how to actually do this efficiently. I think this might be forgivable as I assume many of you are quite young, but are you really suggesting exploiting poor populations for menial work w/o social benefits, without dignity, suggesting they walk into some booth, put in a few hours of work without even the physical presence of co-workers and bosses that can appreciate their work?<p>Again, I'd like to attribute the responses to the commenters' young age, and I trust your good intentions, but hackers, engineers and all entrepreneurs should really learn something about work-relations, social policy and ethics.
"found that you can make about $7/ hour."<p>As someone who turks occasionally, you can't consistently make $7 an hour unless you have some skill that you could probably use on the job marketplace. $3-$4 is more like it.
Mechanical Turk takes up to 30 days to approve payment for any task, so instant payment is out of the question.<p>Having a terminal where someone can sit for hours on end being the same place as a cash dispenser isn't the best idea. The users would probably be better off breaking into it. They need to be separated.<p>Perhaps a better solution would be to start a sweat shop full of terminals where the sweatshop owner pays them on approved tasks, after taking a cut.
I like how everyone's discussing the right and wrong of it, weather it would work or not. A true entreprenuer would ignore all that and just try it anyway. We can talk about this till the world ends but at the end of the day, people much smarter than us have made predictions in all sorts of industries and been outsmarted by 2 guys in a garage who tried something crazy that should not have worked.<p>I say go for it.
I will do him one better. How about he opens a storefront with locked down machines that do nothing but mechanical Turk and rents them out for 1 an hour. Maybe wow gold farming .<p>Anyone have an concerns about the ethics of this? On one had 6 net is better than 0, on the other hand it is fairly predatory.
This is very similar to what Samasource (<a href="http://samasource.org/" rel="nofollow">http://samasource.org/</a>) does except in a different demographic of poverty (in third world countries). The general idea is that in many places, such as refugee camps, no one has any money and there really aren't any jobs to act as segues to help people rise up out of poverty, if you build computer clusters in these areas, they can do mechanical turk style work, make a lot more money than they could otherwise and be supervised such that quality can be more controlled than a bunch of random people on the internet. Unfortunately they run on a different proprietary alternative to Mechanical Turk (presumably with some tweaks to fit their model).
This won't work, the wage rates on Mechanical Turk for unskilled, low motivated users are too low. I previously worked for a microfinance bank in the Philippines where we employed staff at a little over the minimum wage there which was $5 per day. While we were getting ready for a new product launch I had some staff we couldn't use yet, and seeing they all had computers and internet connection ran a small experiment on Mechanical Turk. They averaged $5-10 per day in income so I stopped it after a month.<p>There were several problems:
1. Many of the more valuable task required you to meet certain 'skill' levels which these staff didn't
2. Motivation was a huge problem as spending hours per day clicking on images with people or editing text you don't understand was very dull
3. We had problems with internet bandwidth being too slow to allow running through tasks rapidly<p>It was a pity this experiment failed as it would have been a great way to scale up and create income opportunities for some very destitute people.
For better or worse, you're likely to get better results to your work requests by putting these same devices in chosen offshore locations; you're arbitraging median-skill median-lifestyle foreigners in lower cost of living areas versus lower-quartile lifestyle and limited-skill Americans.<p>Globalization is not friendly to the homeless and mentally ill. This proposal may buoy up some folks who have fallen through the cracks, but there are a number of other social safety nets (a la worker retraining) that offer that same sort of chance to those down on their luck but still able to turn around from a bad situation or bad break.<p>If you are not able to read coherently or quickly - this doesn't necessarily mean uneducated, it could simply be crippling dyslexia - this doesn't help you. And if you are suffering from drug addiction or mental illness or even poor lifestyle choices, this may be significantly less advantageous and doable than panhandling.
This is a non-starter for several reasons, mostly relating to spam.<p>First, you can't just give people cash. They will immediately turn to spamming, and you'll get crap data. You need to do statistics, comparing turks to other turks, and only pay the ones who don't spam. It's highly likely you can't do this in realtime.<p>Second, you need to track the identity of your turks. If worker X is known not to be a spammer, then you want to assign work to X preferentially. Similarly, if worker Y is a known spammer, you want to refuse him work.<p>Third, most poor people are not working and are not looking for work. Only 30% or so of poor adults are in the labor force at all. Why would they decide to start working at a mechanical turk station when they seem to have little desire to work anyplace else?
It's been done (though instead of money, candy and instead of the urban poor, UC Berkeley CS students):<p><a href="http://hci.berkeley.edu/cs260-fall10/images/b/ba/FinalPaper-HeimerlTrinh.pdf" rel="nofollow">http://hci.berkeley.edu/cs260-fall10/images/b/ba/FinalPaper-...</a>
That is actually brilliant! I was thinking about something similar, how maybe a poor country could create jobs for its citizens online (and they'd be happy with much less than $7/hour). But this is actually a great idea for poor people in any country...
This would be a great kickstarter project. You may want to modify it though so the station just prints a check or some other instrument other than cash or you will end up with a lot of broken Mechanical Turk Stations.
Half-baked adaptations:<p>Combine with a Vegas pokie/slot machine. Complete menial hits in exchange for a few more cents to play.<p>Or combine with vending machines. You want that can of Coke? Pay by categorising 30-40 photos quickly.
Be good if Amazon themselves got behind something like this, by providing the banking and payment infrastructure. They already issue credit cards, they could instead issue debit cards and directly pay into those. Would eliminate the need for a cash machine on the terminal, although the absurd ATM fees in the US might make it less viable I guess. I assume there are no charges for using debit cards for small transactions in the US (I live in the UK)?
This sounds like the type of idea that Thiel would be into. I applaud your creativity and these discussions regarding minimum wage are surprisingly pertinent.
What if the person doing the Turking was submitting crap data? It seems like you'd have to wait at least a couple days in order for the data to be validated.
I can see one advantage to this, is that the poor don't have any computer skills, and this may give them basic familiarity with using websites. Although I imagine there would have to be a trainer present at all times to help when they get stuck.<p>A more beneficial concept for the poor is microlending, helping the poor in third-world countries start their own businesses.
The problem with a "solution" like this is same as the problem that the minium wage presents: it is insufficient to live on, and less than the value of social services benefits.<p>If you want to magically improve the plight of the poor, make cheap housing available in urban areas without a qualification process (ie. Section 8, public housing)
What about paywithatweet repurposed so you pay for digital goods with some Turking? Categorise 50 images and get your favourite local band's latest MP3. $1.90 to the band, $0.10 to the intermediary as their cut.
<i>For all the success that micro-lending has had in India and other countries, we haven’t really seen those benefits in American Urban areas.</i><p>I thought the USA had "micro-lending", but they were called "payday loans"?
What this article is effectively advocating, but can't or won't say outright, is that America needs the ability to pay people less than the minimum wage.<p>Someone working all day at mechanical turk, is likely to fall below the federal minimum wage rate in terms of what they're pulling down per hour. Is it ok for companies to utilize mass scale labor at what becomes in reality a sub minimum wage rate? Particularly if mechanical turk stations were to become wide spread.<p>Obviously mechanical turk is a per unit pay system, not a job with an hourly pay rate. However, if you're doing it full time, I call bullshit on that difference. If you had 100,000 people working on mechanical turk 40 hours per week, making $6 per hour, those are very much jobs paying sub minimum wage.<p>It would be no different than if a thousand companies banded together to source labor below minimum wage by paying per task, and sharing that labor around rather than employing each laborer in a "job" (eg in a metro area with high population density). Those companies would be paying for net full time labor, while evading the minimum wage responsibility.<p>One solution to this legal boundary, would be to require that mechanical turk style tasks pay at least equivalent to minimum wage based on the time they take. I expect in any large scale adoption of mechanical turk, this issue will jump to the forefront.
This resonates a lot with Henry Ford's innovations of conveyor belt assembly lines. Ford found out how to humans into an early form of robot, and once technology caught up these jobs inevitably were replaced with actual robots. Even if this Turk Station idea works, how long will it be until AI replaces any need for Mechanical Turk? About 10 years I'm guessing?<p><a href="http://en.wikipedia.org/wiki/Assembly_line#Ford_Motor_Company_.281908-1915.29" rel="nofollow">http://en.wikipedia.org/wiki/Assembly_line#Ford_Motor_Compan...</a>
This is a great idea. It would really work if it supplements the welfare checks; it would get people into the mindset of working. Let me explain.<p>Right now, if you're a bum, you make a decent amount of money panhandling + scamming the government ("disability", etc.); let's call this "freeloading". The problem is: the amount of money you can make freeloading is close to minimum wage (if you include everything). So there isn't much incentive to take up work; freeloading may not pay as much, but you get total freedom, no schedules, no boss, etc. After some time, you just become so used to that lifestyle that it's impossible to come back into the working mainstream.<p>A system like this can be a great way for people to, on their own schedule, supplement their freeloading income. Over time, they'd get used to the concept of effort and reward, and maybe consider taking up real work?<p>On another note: I think minimum wage should be much higher; say, $20/hour. This may not sound like a good idea, but think about it: it should be high enough that a freeloader has serious economic incentive to get off his ass and look for work! Another way to look at it is: I would much rather pay more for something, and have that extra money directly go to a worker's pocket; than route the money circuitously through government taxes, bureaucracy, non-profits, etc. to that person on welfare.<p>Edit: the above are just ideas. If you disagree with them, say something instead of hitting the down arrow.