That its changing rapidly to understand why people are anxious to figure out how concerned they should be. What is changing, what is different this time and what can we do to ensure the coming era of change leads to some degree of shared prosperity and our focus is on the labor market and how we make the labor market function while for as many workers as possible. That is where we see the opportunity and also the challenge, kind of a lesson if you read our report which was just released last week. It is on making the work work and thats why its such an honor to have mary gray from the University Research talking about her book ghost work and here is a photograph. To get a tiny bit of background, her work has been on the sort of communication and use of technology and how they use technology and other means to establish across places and to maintain anonymity. She was recruited to work on projects and once there got interested in the question of how Artificial Intelligence wasnt just something said by workers kind of hidden inside of the machine, so i uniformly misses between machines and people and identities and they are not obvious to the participants involved. They read the book and i thought that it was still quite mind blowing and influenced my own thinking and i was doing. So, we were just delighted to have mary here. Shes going to speak for about 25 minutes about the buck and a fan of the questions so get those ready and will be divided to hear your questions. Thanks very much. [applause] just over a second, how many of you have heard about the recent decision in california, the California Assembly bill number five . I hope it comes up in qanda prime time. How many of you have ever received a 1099 or have done contract work . Wonderful. Wonderful. And how many of you, lets make it a little bit easier to answer, how do you have a friend that its a ride sharing there we go. Can we all have a stake in the conversation. What i want to offer you today is a definition. Its arguably to describe the reorganization perhaps the dismantlement, and we want to think about were. Youll find references to the ghost work and that was quite intentional, and of course anything when you publish it, you dont control it when its out of your hands, so people use that quite a bit. I want to give you an idea of the mechanism behind its workforce. What would want to understand us by the early 2000, computer scientists and engineers realized they could take the same sorts of mechanisms for calling and putting out a particular call through the programming interfaces that would beginning to draw up software to execute a program and spread people into that. So, they quickly ended up creating a mechanism that could help someone with their work which was around the data labeling and vulgar kind of tasks. Its just the beginning of the piece of this you can see, but if we have any business that has a task it wants to demand or request help with. In the abstract, this is going on in this mechanism. So, this is any taskbased project that can at least the scheduled. Like hold on to that and think about how much will th of work s good invalid if you are applying that mechanism to the distribution of the request for work. Thats what i want to suggest. Not just automation, and i think for anyone else, even last year talking about the future of work, it was often like a fear of robots. We started thinking differently. But having to come to the fore with a framework but if it isnt robots, what is it. We are suggesting to you it is potentially ghost work and we need to stop it. You are probably familiar with the Platform Services that we get from the same mechanisms, it is an application interface putting out requests for somebody to come pick up this food, deliver it to this address intand the platform is participating in the exchange by recording when its picked up and delivered into executing a payment, scheduling where it is giving an address. That portion of the work is automated but the delivery and value of somebody being there to give her the food, that is the part of the equation we often are not considering or more increasingly we are aware because they can see. If i said content reservation was a form of the letter i dont think anybody would have known what i was talking about until it came to cambridge and the lupica so you know it is a job people do, absolutely providing another service which is training Artificial Intelligence but there are people performing a very important service. We are focused on this world of business startups and that is the world im going to talk about today. Its the world of editing and some of you in the room might know some of these different tasks do drive Artificial Intelligence innovation, they are what helps structure the clean data set and i would be at mit because most of you know what im saying when i say that but importantly increasingly seeing the number of jobs that are saying it is quite hard we are just going to keep a person into the moment of the Service Request so anything youd gone on to a website and have a little when no popup you know it is a mixture of script and a person assisting user you have a person doing something on the spot that cannot be quite completed by automation is probably a point of reference that comes to mind and we cover this in the book. Its places that are threaded together. It isnt new. Automation will come around. It might look like piecework. Certainly we could have manufacturing knock out a shirt and have a button and go. Usually that was a matter of decades and that is the important thing to take away. Automation eventually made it possible for textiles and labor cost and the reality of the boat itself being too sophisticated for the textile machinery to be able to consume through automation and a person was kept around but its also several contract labor. Women could be brought in at the time to service computers and when they need and deman the ner them eventually disappear, they could be let go. There was no security in that employment, it wasnt as valuable or there wasnt a way of looking at employment and seeing these women is valuable precisely because they have already moved forward. Often this particular buddy is embodied, white and of privilege who have specific goals to play so anyone not playing those roles seemed expendable so to continue the lineage, by the 1960s and the advancement of staffing and Temp Services that quite literally was brokered on the value of womens labor as a resource because these were young women that were College Educated and make great office girls. But they also were expendable. So, keeping that thread moving by the time we get to the 1980s and 2,000 in the outsourcing and off shoring of knowledge work, it becomes much harder to make the case people are doing something that could so easily be replaced precisely because they are doing work also being done by workers in the united states. It becomes much more obvious it is a question of labor arbitrage you can find cheap labor that is just as educated as anyone who ran the location of generating the request for work. And ive often lamented perhaps as some of my colleagues a settlement of the case against microsoft or them they never resolved the question of what do you do in the case of employment that is necessary for a period of time and it might be something you need somebody with a specific kind of expertise, language expertise, coding expertise that you know you are not going to need them for more than 12 for 12 weeks or 12 days, what ways do we have to value that worker. Its important to Silicon Valley in 2001 what happened we had an effectively the dot com bubble burst and without the case law it was going to be questionable to do with people who were necessary but not necessary as if we are going to hold them for a career and we came up with a settlement a set of practices that through Management Systems that dont leave them with the protection beyond that contract to be able to say i am employed in these benefits come with my employment. So if you think about the history that weve drawn and its an argument in any trajectory of argument that in this case we see the beginning of the industrial era of them all around me were per protection that we see the valuable work is the work that cant be automated without much projection out. So the policies both for the Assembly Line and professions that were imagined to be beyond the touch of automation and growth of the Service Industry that served peoples request for needs more than it serves the need to build something. Lastly, to see the shift towards arguably the information economy that involved a lot of people doing Information Service work. Guess it involved people doing coding and other really valuable skills to take a great amount of training. But if we think about what it took to code a website, how many people in the room know what im talking about, thats now completely done with software. At that time we paid quite a bit of money to build this website. That was some of my first work. Keep that in mind when you think about what can be automated and what is the work of being an automator the creative work, complex communication that goes beyond automation, it is an open question. What is it that will be constantly on the horizon but we cant see it growing into a career. Let me give you how we studied this world because studying a distributed world of work is not the most obvious thing to do. It was hard to figure out where to dig in. We chose for businesses and also these companies to show us the inside of their blackbox how to be organized, what does the workflow look like and the companies that stand in for the mechanism that can both build out Artificial Intelligence and also keep humans in the loop. This two streams of work and what i hope you take away is to see the growing work is really growing into different directions. One is the increasing need to structure and analyze data and so take the example of the radiologist. If it isnt a one and done matter if you are able to do a great job getting Image Recognition for particular problems in radiology that seem like theyre going to get rid of the radiologist with the other market phenomena are happening you expanded the market for people using radiology you can create mechanisms getting outside urban centers and places no one has ever been able to get access to. Weve also created a market for other kinds of diagnostics that are going to require an amount of domain expertise by a medical professional so its not a doctor doing the work we radiologist. It might be a medical student oriented professional on the horizon. Its managing the information into thand the data they are cog that will build and take over the need for the human hand in the mix. But then the Second String which is just as critical, human and other Information Services where we come to expect as consumers theres some one to answer my call 24 seven. I would love if we didnt have ththat expectation to feel like someone is expecting that of me. Take your medication if you are taking care of an elderly parent and wants not only that but something to tell you how is the care going to take him and these kind of cases or the role of humans in the loop that are not fulltime employment, their task driven work. In the case of the studies they are analyze analyzing the case f companies that both do image tagging verification if weve picked these things people thought these are easy to automate and it turns out they are not so easy to automate