I don't disagree with this sentiment, but I don't think it is correct to say that Mechanical Turk itself is abusive. Mechanical Turk certainly CAN be abused, but it also CAN be used ethically. For example, we regularly use MT to complete tasks that pay workers effective hourly wages in the range of $8-20 per hour, and I know several other requesters who pay similarly.
It is probably most accurate to say that the lack of worker safeguards on MT attracts many users who pay rates that are (IMHO correctly) classified as abusive. Part of the issue with MT is that traditional worker protections like a minimum wage no longer make sense when you have compensation that is paid on a per-task rather than per-hour basis. Under a per-task compensatory system, faster workers will always earn more money (on an hourly basis) than slower workers. For the tasks we do, it is not uncommon for subsets of workers to earn effective hourly rates almost three times as high as others. In many ways this is good - incenting workers to spend their time on tasks that they can complete most effeciently (and away from tasks that they are less suited for) helps promote the most efficient allocation of labor resources.
But how do you ensure that all workers will always be treated fairly when a minimum wage no longer makes sense? What do you replace it with? Its not clear that there is a simple replacement that works in the micro-tasking context.
Our platform (http://houdiniapp.com) helps businesses complete projects on MTurk. One of the things we realized is that many requesters pay low wages simply because they are really bad at estimating how long tasks actually take to complete. We’ve been able to solve this (somewhat) by providing tools that provide real-time feedback on average completion times in order to provide direct visibility into real hourly wages. However we allow users to set their own compensation and some users still end up paying effective wages in the $2-3 (which we do not encourage, but don't have any simple way to prevent).
What makes you think that Facebook allowed anything? What power does Facebook have to dictate anything to Instagram prior to purchase?
I think you have a misunderstanding of how deals like this take place. Facebook would have undoubtedly preferred to close the sale prior to the B round, but I am sure that whenever Facebook originally came calling, Instagram maintained that they were not interested in selling. I know this only because this is what hot target companies always say when receiving unsolicited offers. Instagram followed up by putting its money where its mouth was and closing a big round that gave it enough money to stay independent for a long time while growing the company. At that point, Facebook's only option was to go big or go home.
Facebook could have tried to preempt the B-round by raising its bid before the round closed, but it seems pretty clear that it was going to take about $1B to get Instagram off the block, whether that happened before or after the B round.
At the end of the day, the B-round does not make a huge deal to Facebook either way - to the extent that you are prepared to pay $1B for a company, you're not going to be too worried about a measly $50M going out the door.
Its actually not that surprising upon closer inspection. The articles that were focused more general topics (The Cost of Cracking, How to Make it on Craigslist) were generally correlated, while the post that were more hacker specific (Minimum Viable SEO, Headphones for Hackers) did well on HN, but did not receive a lot of general traffic.
Very interesting post. How has the blog impacted your user growth? Given the effort involved in writing each post, has it been worth it in terms of growing the business?
You seemed to have skipped right over the first few sentences from the section you cited: "Nolan confirmed that the ambiguity was deliberate, saying "I've been asked the question more times than I've ever been asked any other question about any other film I've made... What's funny to me is that people really do expect me to answer it....I put that cut there at the end, imposing an ambiguity from outside the film."
:)
I agree that resolving the entire film as simply a dream would be a huge let down. If it were another film, would be content to stop there. For my money though, resolving that the ending puts Cobb firmly back into reality is also highly unsatisfactory. The more interesting (and I believe) intended result is force the viewer to question their own sense of reality. Haven't you ever had a dream that was so realistic that you were certain it was real - until you woke up? How do know for sure that we are not simply living a dream that we will one day wake up from?
There is a book by Stanislaw Lem that beautifully explores this sort of idea. In the story, there is mad scientist fellow that has a room full of electronic brains, each being slowly fed a life story via a series of slowly rotating magnetic drums. To the individual brains, the story that they are being told IS their life - they have no idea that they are simply boxes in some mad scientist's laboratory. The book goes onto suggest that our own lives may simply be programmed by a mad scientist who exists a level up from us.
I believe this it is this sort questioning of reality that Nolan is trying to impress upon the viewer. Reducing the whole thing to just a sci-fi film about a couple of dreamwalkers makes it seem frankly one-dimensional and uninteresting.
Its a fantastic app - I've been using it for about two weeks and its completely replaced a clunky process of trying to link lead spreadsheets, with Highrise contacts and Gmail emails. We store contacts in Highrise but it is pretty lousy for pipeline management. Because Streak lives inside gmail, it makes it really easy to see all the email associated with each lead. It also replaces Boomerang (email scheduler applet) and Tout (email template app) so I can get more done without leaving my inbox. Streak and Rapportive handle almost all my CRM needs at this point.
This seems to be an overly simplified take. If Nolan had intended to make it clear that the final scene was indeed reality then he could have just as well shown the totem falling while Cobb walked away. It seems pretty clear that Nolan explicitly intended to leave it up to the viewer to decide what was "real" and what as not. This ambiguity between dream and reality is one of the most important themes in the film - Cobb spends the entire movie preoccupied with keeping track of reality, to the point that one could argue that Cobb's had been incepted to remain obsessed with identifying reality to the same degree that Mal had been incepted to perpetually believe that she was dreaming. When Cobb sees Mal in limbo at the end, she also makes point explicit - pointing out that Cobb has simply chosen to believe that his children "up there" are what is real and that Mal "down there" is not.
The final scene preserves this ambiguity, while underscoring the fact that the obsession with reality is no longer important to Cobb - he is finally at peace with where he is - real or not.
Even if he suffered a loss (i.e. if the appreciation in the home value was not enough to offset his accumulated outflow of principal and interest) it probably would not be as much as he would have lost by renting.
Your scenario assumes that (1) he could rent the same or an equivalent property for much less of cash outflow than buying and (2) he could have invested the difference in some other asset that had a much higher return than appreciation of the condo.
If the amount the rental amount is about the same amount as the mortage payments then you are almost certainly going to be worse off renting than buying under average market conditions.
This is great. I have no idea whether this idea will work or not, but I really enjoyed reading about your process. It sounds like you have really embraced the lean startup, customer development philosophy. I have to admit that I had never heard of you until I saw you interview on Gabriel's website, but have been intrigued ever since. Thanks for sharing.
It is probably most accurate to say that the lack of worker safeguards on MT attracts many users who pay rates that are (IMHO correctly) classified as abusive. Part of the issue with MT is that traditional worker protections like a minimum wage no longer make sense when you have compensation that is paid on a per-task rather than per-hour basis. Under a per-task compensatory system, faster workers will always earn more money (on an hourly basis) than slower workers. For the tasks we do, it is not uncommon for subsets of workers to earn effective hourly rates almost three times as high as others. In many ways this is good - incenting workers to spend their time on tasks that they can complete most effeciently (and away from tasks that they are less suited for) helps promote the most efficient allocation of labor resources.
But how do you ensure that all workers will always be treated fairly when a minimum wage no longer makes sense? What do you replace it with? Its not clear that there is a simple replacement that works in the micro-tasking context.
Our platform (http://houdiniapp.com) helps businesses complete projects on MTurk. One of the things we realized is that many requesters pay low wages simply because they are really bad at estimating how long tasks actually take to complete. We’ve been able to solve this (somewhat) by providing tools that provide real-time feedback on average completion times in order to provide direct visibility into real hourly wages. However we allow users to set their own compensation and some users still end up paying effective wages in the $2-3 (which we do not encourage, but don't have any simple way to prevent).