Return to Video

Databite No. 119: Mary L. Gray

  • 0:00 - 0:09
    [music playing]
  • 0:09 - 0:12
    Hi everybody. Welcome to Data and Society.
  • 0:12 - 0:14
    My name is Sareeta Amrute
  • 0:14 - 0:16
    I'm the director of research here.
  • 0:16 - 0:19
    It's my sincere pleasure and honor
  • 0:19 - 0:22
    to welcome you to Data and Society
    for this discussion
  • 0:22 - 0:26
    inspired by Mary Gray and
    Siddharth Suri's recently released book
  • 0:26 - 0:32
    Ghost Work: how to Stop Silicon Valley
    from Building a New Global Underclass.
  • 0:32 - 0:36
    Mary Gray is senior researcher
    at Microsoft Research
  • 0:36 - 0:41
    and fellow at Harvard University's Berkman
    Klein Center for Internet and Society
  • 0:41 - 0:44
    Mary also maintains a faculty position in
  • 0:44 - 0:47
    the School of Informatics Computing
    and Engineering
  • 0:47 - 0:52
    with affiliations in anthropology,
    gender studies and the media
  • 0:52 - 0:54
    school at Indiana University.
  • 0:54 - 0:58
    Her research looks at how technology
    access, material conditions,
  • 0:58 - 1:02
    and everyday uses of media
    transform people's lives
  • 1:02 - 1:04
    and today she'll be talking to us
    about her latest book
  • 1:04 - 1:07
    written with Siddharth Suri
    who is based in Seattle.
  • 1:08 - 1:09
    Take it away Mary.
  • 1:09 - 1:15
    [applause]
  • 1:15 - 1:17
    Thank you.
  • 1:17 - 1:18
    Thank you everyone for coming out
  • 1:18 - 1:22
    and I see some familiar faces
    and I just really want
  • 1:22 - 1:24
    to voice my appreciation
    for all the support I've had
  • 1:24 - 1:26
    over the years doing this work
  • 1:26 - 1:29
    first and foremost
    to my co-author Sid Suri
  • 1:29 - 1:31
    but to all the workers
    who have given their time
  • 1:31 - 1:35
    and let us into their lives
    to learn about their experiences.
  • 1:35 - 1:38
    This work wouldn't be possible without
    the time that they've given to us.
  • 1:38 - 1:40
    So with that
  • 1:40 - 1:45
    I wanted to start by giving you a sense
    of where this work came from
  • 1:45 - 1:49
    and for me, I was thinking
    about my own research questions
  • 1:49 - 1:51
    before coming to Microsoft Research.
  • 1:51 - 1:54
    Most of them circled
    around the question of
  • 1:54 - 1:58
    how do we become more or less seen?
  • 1:58 - 2:03
    How are we known and valued as people and
    what role do technologies play in that and
  • 2:03 - 2:08
    much of the heralding of the internet was
    that we're going to become more visible.
  • 2:08 - 2:11
    We're going to be able
    to speak our truth, hear all voices
  • 2:11 - 2:15
    and much of my grounding in anthropology
    and critical media studies
  • 2:15 - 2:18
    brings me to the question of ''how so''?
  • 2:18 - 2:19
    ''When is that not true''?
  • 2:19 - 2:24
    and ''what are the conditions under which
    people make that more or less true''?
  • 2:24 - 2:28
    So I come to this project
    with that background.
  • 2:28 - 2:30
    And in many ways
  • 2:30 - 2:34
    what I'm hoping to do is incite you to
    care about this world of work
  • 2:34 - 2:38
    that is more or less seen,
    known and valued
  • 2:38 - 2:41
    depending on where you are
    in this universe.
  • 2:41 - 2:45
    It really started with coming
    to Microsoft Research
  • 2:45 - 2:49
    and asking a basic question about
    how artificial intelligence is made.
  • 2:49 - 2:51
    I had no idea
  • 2:51 - 2:54
    and so when I started asking computer
    scientists and engineers in my lab
  • 2:54 - 2:58
    what goes into developing
    algorithms and the models that are built
  • 2:58 - 3:00
    to be able to advance
    artificial intelligence.
  • 3:00 - 3:04
    It turns out that there are a lot of
    people involved in that work
  • 3:04 - 3:07
    outside of the coders and the engineers
    and computer scientists that are
  • 3:07 - 3:11
    theorizing these, these
    technological innovations.
  • 3:11 - 3:16
    It's a lot of people who are
    effectively cleaning and managing data
  • 3:16 - 3:21
    the training data that become the models
    for building algorithms out
  • 3:21 - 3:25
    and there isn't a case of any
    artificial intelligence that exists
  • 3:25 - 3:30
    that doesn't depend at some point in
    someone touching that data
  • 3:30 - 3:34
    curating that data and taking something
    that's otherwise kind of
  • 3:34 - 3:39
    structuralist nonsense and putting it
    into some structured sense
  • 3:39 - 3:43
    that a computational process
    could then model and learn from.
  • 3:43 - 3:47
    So at the goal of this book, if there's
    nothing else you take from this book,
  • 3:47 - 3:52
    it's to understand that artificial
    intelligence always has human hands in it
  • 3:52 - 3:56
    that we are benefiting from a lot
    of people contributing to advancing these
  • 3:56 - 4:02
    technologies, even in cases where we might
    fully automate one process along the way
  • 4:02 - 4:06
    particularly as its impact
  • 4:06 - 4:10
    or its application to a domain
    it wasn't expected to enter
  • 4:10 - 4:14
    say language, like
    text translation that's done in real time.
  • 4:14 - 4:18
    If you're a speaker of multiple languages
    and you're code switching
  • 4:18 - 4:21
    odds are pretty good that the AI isn't
    going to be able to keep up with you.
  • 4:22 - 4:26
    So look at those cases where you then
    have to bring people back into the mix
  • 4:26 - 4:30
    to be able to develop a model that
    would be able to capture
  • 4:30 - 4:32
    what kind of exchange
    is happening.
  • 4:32 - 4:36
    That's really the beginning of this book
    is to understand who are the people
  • 4:36 - 4:40
    who are doing all of this work and
    it turns out that when you ask
  • 4:40 - 4:46
    computer scientists and engineers often
    their responses are I don't really know.
  • 4:46 - 4:50
    I've never really met these people.
    The beauty of this technology is that
  • 4:50 - 4:55
    I don't have to meet them and
    I say that now with all seriousness
  • 4:55 - 5:00
    the sense is that this is a technological
    innovation often called human computation
  • 5:00 - 5:05
    or crowdsourcing the technique of being
    able to thread a person into a moment
  • 5:05 - 5:10
    of judgment where you need a person
    to be able to evaluate or decide something
  • 5:10 - 5:13
    that a computational process
    can't quite figure out.
  • 5:13 - 5:18
    Bring that person into that moment, that
    judgment and then threading them into
  • 5:18 - 5:21
    a computational process, an automated
    process so that you can carry on
  • 5:21 - 5:23
    with an output.
  • 5:23 - 5:27
    So we're somewhat familiar with some of
    the applications they're little bit
  • 5:27 - 5:29
    visible to you today.
  • 5:29 - 5:30
    This is an iceberg.
  • 5:30 - 5:33
    Most of you are familiar
    perhaps with your Uber driver.
  • 5:33 - 5:36
    You've met them
    maybe you've chatted with them.
  • 5:36 - 5:39
    You might be familiar
    with other platform services
  • 5:39 - 5:42
    on demand services
    that effectively
  • 5:42 - 5:44
    are using the same technology
    of human computation
  • 5:45 - 5:48
    to match a person
    who's able to deliver a service
  • 5:48 - 5:51
    through a mix of
    application programming interfaces
  • 5:51 - 5:53
    that calls that person to the job
  • 5:53 - 5:55
    whatever it might be
  • 5:55 - 5:57
    whether it's to pick you up at the airport
  • 5:57 - 5:59
    or to pick up some food
    and bring it to your door
  • 5:59 - 6:01
    or if it's a content moderator
  • 6:02 - 6:04
    And I think what's fascinating
    is two years ago
  • 6:04 - 6:07
    If I'd said the phrase content moderator
    or content moderation
  • 6:07 - 6:10
    I would have just gotten blank looks
  • 6:10 - 6:13
    How many of you know
    what content moderation is today?
  • 6:13 - 6:17
    It turns out they're doing
    an incredibly important job.
  • 6:17 - 6:19
    They are people who effectively curate
  • 6:19 - 6:22
    look at pieces of text and images
  • 6:22 - 6:25
    that are beyond the capacity
    of any computational process
  • 6:25 - 6:31
    to analyze and evaluate and say:
    ah, that's pornography or that spam
  • 6:31 - 6:33
    versus someone sharing information.
  • 6:34 - 6:37
    So it turns out that
  • 6:37 - 6:40
    we're not that far along
    in being able to evaluate text and images
  • 6:40 - 6:44
    to figure out is that content
    that should or should not be there.
  • 6:44 - 6:48
    You could say anything that's hard
    for a human to evaluate and decide
  • 6:48 - 6:50
    is that misinformation?
  • 6:50 - 6:55
    or just a fact
    that I'm not familiar with yet?
  • 6:55 - 6:56
    The odds are very good
  • 6:56 - 7:00
    a computational process isn't even close
    to being able to figure that out.
  • 7:00 - 7:02
    If it's hard for a person to figure out
  • 7:02 - 7:06
    it's going to be
    intractably a technically hard process
  • 7:06 - 7:09
    for computation to model.
  • 7:09 - 7:14
    You have to have
    really certain this or that, yes or no
  • 7:14 - 7:19
    to build code with accuracy;
    to be able to automate something.
  • 7:19 - 7:22
    So again, take away how much this world
  • 7:22 - 7:26
    that's completely dependent
    on having people at a moment of judgment
  • 7:26 - 7:31
    enter the scene like content moderation
    and then look below that surface
  • 7:31 - 7:34
    and that surface below that iceberg
  • 7:34 - 7:40
    that is this spiraling, growing,
    expansive world of services
  • 7:40 - 7:45
    that effectively are building
    to keep a person in a computational loop
  • 7:45 - 7:48
    because it turns out it's
    much more efficient and effective
  • 7:48 - 7:53
    to be able to match a person
    to a task like captioning and translation
  • 7:53 - 7:57
    or a task that might be image tagging
    for a new set of images
  • 7:57 - 8:01
    that you're trying to evaluate,
    whether it's for training AI
  • 8:01 - 8:04
    or that you want to do
    a marketing project
  • 8:04 - 8:06
    in all of those cases,
    all of these businesses
  • 8:06 - 8:11
    that are probably unfamiliar to you
    that are on this on this slide
  • 8:11 - 8:14
    are quickly making the best
    of a business model
  • 8:14 - 8:18
    that brings contract driven,
    task-oriented work
  • 8:18 - 8:21
    to people mostly doing work in their homes
  • 8:21 - 8:23
    or if they're in a setting
  • 8:23 - 8:26
    they're covered by what are called
    vendor Management systems.
  • 8:26 - 8:30
    And again, they're people that
    you will never meet as an end consumer
  • 8:30 - 8:33
    but that you benefit from every day.
  • 8:33 - 8:36
    So when I'm asking engineers
    and computer scientists
  • 8:36 - 8:41
    about this work of human computation
    and the role of people in the loop,
  • 8:41 - 8:44
    it turns out that most of these businesses
  • 8:44 - 8:47
    are effectively doing
    what these Engineers are doing,
  • 8:47 - 8:49
    which is bringing people in
    as quickly as they can
  • 8:49 - 8:52
    and then moving on to the next project.
  • 8:52 - 8:54
    They're not asking "Who are these people?
  • 8:54 - 8:57
    Under what conditions
    might they be working?"
  • 8:57 - 9:03
    In most cases they're working on contract
    for that specific task itself.
  • 9:03 - 9:05
    So the moment of engagement
    might not last
  • 9:05 - 9:08
    more than a few minutes at best.
  • 9:08 - 9:11
    So it's a pretty kaleidoscopic world.
  • 9:11 - 9:15
    So at Microsoft research
    I feel incredibly lucky
  • 9:15 - 9:16
    to be around people
  • 9:16 - 9:19
    who do reflect on this question
    of what are they building
  • 9:19 - 9:21
    for the rest of the world.
  • 9:21 - 9:23
    And in many cases
    when I meet a group of people
  • 9:23 - 9:28
    who are saying "I don't really know
    who are the workers who are here."
  • 9:28 - 9:30
    There's at least a subset of those folks
    who will say,
  • 9:30 - 9:32
    " I don't exactly know
  • 9:32 - 9:35
    but you know, the technology
    really keeps me at a distance".
  • 9:35 - 9:39
    and then there was a third set
    that would answer fairly regularly
  • 9:39 - 9:43
    "I don't know
    and I don't know if I want to know".
  • 9:43 - 9:46
    and as you can imagine
    for any anthropologists in the room,
  • 9:46 - 9:49
    that's just, that's...
    you really want to pursue that question.
  • 9:49 - 9:51
    What makes somebody uncomfortable
  • 9:51 - 9:55
    about knowing who is on
    the other side of a screen?
  • 9:55 - 10:00
    What makes it seem an intractably
    socially uncomfortable question
  • 10:00 - 10:03
    to find out about their work conditions?
  • 10:03 - 10:04
    So when I met Sid Suri,
  • 10:04 - 10:07
    he was really one of the first people
  • 10:07 - 10:10
    who genuinely,
    coming out of computer science,
  • 10:10 - 10:12
    wanted to not only know
  • 10:12 - 10:14
    what work conditions
    people might be engaging in
  • 10:14 - 10:16
    but what their lives were like.
  • 10:16 - 10:21
    And so we started on this journey
    and I don't use that word lightly.
  • 10:21 - 10:24
    It took us five years
    to develop a methodology
  • 10:24 - 10:30
    for being able to bring
    the value of qualitative critical work
  • 10:30 - 10:32
    that engages people
    in their everyday lives
  • 10:32 - 10:33
    and figure out
  • 10:33 - 10:36
    where you could integrate measurement
    and computational analyses
  • 10:36 - 10:40
    to build out
    a picture of this world of work.
  • 10:40 - 10:44
    So, often we'll get this question of,
    "Well, how big is this Market?"
  • 10:44 - 10:47
    Underneath that question
    is often people who feel like
  • 10:47 - 10:49

    "Why should we bother caring?
  • 10:49 - 10:51
    This is work
    that's going to by automated any day."
  • 10:51 - 10:56
    If you buy the beginning premise
    of this book, and I hope you do,
  • 10:56 - 10:58
    this work isn't going away.
  • 10:58 - 11:01
    The tasks will change,
    but in fact we're building towards
  • 11:01 - 11:06
    a world of a service industry
    Information Services knowledge work.
  • 11:06 - 11:08
    That isn't a niche job.
  • 11:08 - 11:12
    This is the dismantlement
    of full-time employment.
  • 11:12 - 11:17
    The dismantlement of full-time employment
    for anyone who does creative work.
  • 11:17 - 11:21
    So, we might not be able to see
    how large that market is.
  • 11:21 - 11:23
    Arguably, we've never done a head count.
  • 11:24 - 11:27
    There is no effective way
    to do a worker census
  • 11:27 - 11:31
    of an environment that is by design
    distributed, global,
  • 11:31 - 11:36
    and often doesn't have a category of work
    that people would recognize
  • 11:36 - 11:41
    and resonate with, where they could say
    "Yeah, that's me. I do that work."
  • 11:41 - 12:06
    So this is both a world in which
    our old categories of what job do you
  • 12:06 - 12:06
    do is being blown apart that's
    going into this world of work
  • 12:06 - 12:07
    and it's a world in which we
    don't have any mechanisms
  • 12:07 - 12:07
    for tracking and holding
    accountable the supply chain
  • 12:07 - 12:07
    That's going into
    this world of work.
  • 12:07 - 12:07
    but let's take some guesses here
    and some of this is drawing
  • 12:07 - 12:08
    on economics and the secondary
    literature about the possible size
  • 12:08 - 12:11
    of this market right now.
    We know that there's about an estimate of
  • 12:11 - 12:14
    About five percent of the U.S.
    population alone.
  • 12:14 - 12:20
    According to Pew that's doing
    some form of online work
  • 12:20 - 12:21
    so that there are at least in part
    the work is sourced, scheduled
  • 12:21 - 12:34
    managed, shipped and build
    through an application programming
  • 12:34 - 12:35
    interface an API and the internet,
    five percent that doesn't necessarily
  • 12:35 - 12:35
    mean they're doing and their entire job online.
  • 12:35 - 12:40
    It means that a form of income that's important to them is coming from one of these jobs.
  • 12:40 - 12:49
    Now, this is really striking if you can take into consideration, we've only had the possibility of making an income from this form of work for about a decade.
  • 12:49 - 12:51
    So to have five percent of the u.s.
  • 12:51 - 13:09
    Population already doing this work start thinking through the size of this Market the growth of this market and if that's not quite enough think about how large the global market for the businesses generating value will be by next year.
  • 13:09 - 13:11
    It's a twenty five billion dollars.
  • 13:11 - 13:19
    The street already and that's not a small number if you think about how it compares to other industries that are fairly mundane.
  • 13:19 - 13:52
    So that iceberg that I showed you all of those businesses that are spinning up on below the surface of our visibility as consumers is building an incredible amount of economic value that up to this point doesn't seem to be moving to the other side of the screen to the workers themselves who are doing the work another projection here of thinking about the implications of dismantling employment is to imagine this is not the displacement or the full automation of occupations and work.
  • 13:53 - 14:03
    It's the semi automation of work and being able to task a fi it that is the target of most of these industries and the technologies that companies are building.
  • 14:03 - 14:11
    Everybody would like to figure out how to break down things like scheduling managing any sorts of appointments any of your work.
  • 14:11 - 14:29
    Low figure out how to break that down and turn it into a task that you can hand off to someone else so that you can up whatever it is that is your main point of view or value right that this is again the object of most of the industry on building out the Technologies at the rate.
  • 14:29 - 14:33
    We're going we're looking at 38% of jobs within the u.s.
  • 14:33 - 14:37
    Moving to being semi-automated by the 2030s.
  • 14:37 - 14:52
    Now that might sound like a shocking number but let's just make it mundane that means taking most of office work knowledge work Information Services and turning it into contract work that's already happened in a lot of places.
  • 14:52 - 14:55
    So it's not as though this is so futuristic.
  • 14:55 - 15:00
    And in many ways this the goal of this book is to say it is not too late to
  • 15:00 - 15:05
    Mine what this world looks like we're really just at the beginning but know that it's moving quickly.
  • 15:06 - 15:09
    So the kinds of work
    that this entails again
  • 15:09 - 15:12
    because it's because you can't see it.
  • 15:12 - 15:14
    It's often really hard to describe.
  • 15:14 - 15:19
    It's everything from editing copy,
    editing content curation.
  • 15:19 - 15:38
    The should be if anybody participates in these kinds of jobs, you might see yourselves in these these tasks taking surveys marketing design any sort of graphic design any sort of data entry and labeling which is a pretty labor-intensive cognitively hard job.
  • 15:38 - 15:49
    If you're constantly looking at again data that's coming in that scraped from somewhere with not a lot of context and you're trying to figure out what would you call this material?
  • 15:50 - 15:53
    Is this analyzing somebody's attitude about a product.
  • 15:53 - 15:57
    How would you how would you assess that attitude?
  • 15:57 - 16:03
    It's actually pretty challenging work and it goes very quickly kind of task by task.
  • 16:04 - 16:06
    We were studying very specific.
  • 16:06 - 16:26
    He's I'm going to just go through a bit of the methodology that we used but we were looking at companies that generate sales leads so you can scrape the web and get an idea of who your contact person might be if you sell air conditioners, but you're going to do much better doing your sales of you know, who you should call in that office turns out generating sales lead.
  • 16:26 - 16:46
    That's a particular kind of on-demand ghost work being able to take what is otherwise just a web scrape of people's contact information and curating that list and figuring out who should I contact and then handing it over to a business that wants the best contacts that's a very specific vertical within this industry.
  • 16:46 - 16:52
    And then translation Ted has one of the largest projects open translation projects.
  • 16:52 - 17:06
    That was the beginning of a volunteer community invested in making videos available for hard of hearing communities and for linguistic diversity, and it was the the heart and soul.
  • 17:06 - 17:22
    Amara.org which is one of the companies that we one of the organizations that we studied and then the other kinds of tasks that again are becoming more familiar to some of you content moderate moderation classification tasks that are meant to optimize your search query experience.
  • 17:22 - 17:26
    So if you're typing in something and this will happen happen every election year.
  • 17:26 - 17:30
    If you have a new candidate up for election odds are pretty good.
  • 17:30 - 17:44
    If you've never seen that candidate before they had to do some work to make sure that when people were searching that term searching that name that it matched to the proper biography or persons official presence online.
  • 17:45 - 17:51
    So it's it's kind of this renewal of a need for making sure the information is relevant.
  • 17:51 - 18:06
    I love the example of how many of you remember a moment during the past elections when Romney made a reference to binders full of women right to be able to figure out should that be a trending topic because if you just think about that phrase that's a nonsensical phrase.
  • 18:07 - 18:16
    Aside from who said it and so realize that took a lot of content moderators working very quickly to be able to identify.
  • 18:16 - 18:17
    Oh the context for that.
  • 18:18 - 18:21
    Oh, it's an election cycle a candidate debate.
  • 18:21 - 18:25
    Yes, trending makes sense that's happening below the surface.
  • 18:25 - 18:37
    You'll never see it happen and it's not something that can be automated anytime soon as would be my argument and then lastly thinking about these mundane uses of location verification.
  • 18:37 - 18:40
    How many of you had a favorite restaurant that went out of business.
  • 18:40 - 18:53
    Last month odds are If It Moves somewhere the location needed be re verified and updated within search queries that's still very much the handwork of people below the API.
  • 18:54 - 18:56
    So there's a lineage here.
  • 18:56 - 19:03
    This is not so new and and we take great pains to point out that the tendency to treat contingent work.
  • 19:03 - 19:06
    That seems like it's going to go away anytime.
  • 19:06 - 19:14
    Therefore not that valuable and something we don't need to care about in terms of our employment relationships.
  • 19:14 - 19:48
    That's old news. So if you think about the experience of the Industrial Age and peace work and the work that was literally something that couldn't be accomplished by the newly quickly moving Loom, but that could be taken to mostly Family Farms and in the United States context and be able to share that material that raw material and the material that's been created say a shirt and to have the button or the flourish of bows that could then be attached by a person.
  • 19:48 - 19:53
    That's the kind of work that the entire time the Hope was eventually that will be automated away.
  • 19:54 - 20:06
    And yes, eventually the machines were able to attach the flourish the button the bows that didn't mean that the work entirely displaced other kinds of work that needed to be put on the table.
  • 20:06 - 20:11
    This could go on for generations and did in manufacturing arguably.
  • 20:11 - 20:23
    The only reason automation can knock it out of the park is precisely because you can build the factory around the automated mechanical processes and get people entirely out of the building.
  • 20:23 - 20:38
    But in any case where you're working with people and effectively when you're trying to serve their interests and anticipate their needs you're in an entirely different world of required tasks on a person's time and cognitive ability.
  • 20:39 - 20:57
    So if you think about the next generation of lineage here and the computers behind being able to put people in space or to be able to do some amazing technical achievements, we've always had these moments where the Assumption was.
  • 20:57 - 21:01
    There's something wrote and on Creative about this work.
  • 21:01 - 21:06
    It can be done by an often is done by the same suspects generation.
  • 21:06 - 21:16
    Generation, but their work is not seen as integral to what is really valued and worth retaining or underwriting through full-time employment.
  • 21:16 - 21:28
    So for much of the women who are involved in the Cold War projects through NASA and through other Aeronautics institutions, they were on contract.
  • 21:29 - 21:31
    They didn't have full-time jobs.
  • 21:31 - 21:44
    They could be they could be released at any point and particularly before it was illegal to fire women because they were pregnant women could be dismissed as soon as they married because odds were good.
  • 21:44 - 21:45
    They would get pregnant.
  • 21:46 - 22:05
    So thinking about this lineage is important to imagine who is often invited to take on these contract positions precisely because into the 60s there imagine to be the perfect temporary Workforce both able to do the work as in Kelly girls and selling Kelly girls as the
  • 22:06 - 22:19
    Opportunity for business professionals to have someone take care of their needs and then quickly exit to bring in fresh Minds fresh bodies for the work that needs to be done around the office.
  • 22:20 - 22:58
    So you might see a pattern developing here in this lineage of who is seen as replaceable or less valuable and therefore ripe for contract work as we move into the 90s late 90s to early 80s to late 90s and the internet and connected communication devices allow the workflow of office work that otherwise seemed the domain of professionals from accounting to human resources to any sort of financial services.
  • 22:58 - 23:06
    It becomes quite easy to take that work and move it to other continents where you have enough linguistic capacity to be able to
  • 23:06 - 23:12
    Take advantage of Labor Arbitrage a cheaper Workforce and still be able to get the work done.
  • 23:13 - 23:25
    So what I'm hoping you see in this lineage is precisely the set of assumptions that say who's not so valuable here, who is the person that should hold this temporary job?
  • 23:25 - 23:47
    Because we don't necessarily need to care about them too much and what are all of the ways in which contingent work particularly the United States sets in motion a framing of contract work as disposable less important contingency becomes a value proposition to the business not to the worker so
  • 23:49 - 23:54
    The way in which we went about studying this and I can go into this in the the QA.
  • 23:55 - 24:02
    It's really hard to figure out how to find people behind a distributed system who are working globally in their homes.
  • 24:02 - 24:18
    I'll be the first to confess that I'm someone who really likes to find the people that people assume or otherwise really hard to find because it turns out if you just ask people hey, where are those people they'll prickly quickly identify themselves and say, oh I do this work.
  • 24:18 - 24:20
    This was a whole other level of Challenge.
  • 24:20 - 24:34
    And so the way we went about it was to find four institutions organizations that were producing this kind of Labor and then to figure out ways in which we could meet the workers who were participating in these labor markets.
  • 24:35 - 24:47
    We work agnostically with the assumption that this is contract work unless it's otherwise called something else and we worked with the terms of the of the people who are engaging in these projects.
  • 24:48 - 24:48
    So we studied
  • 24:49 - 25:02
    So Mechanical Turk, which set the Baseline for how most of task-based work is framed and treated and the sidestepping of any legal Frameworks or classification that go with it.
  • 25:02 - 25:26
    And then we also looked at the universal human relevant system, which is the internal platform that Microsoft has please note every large tech company has an internal platform that you can't see that's larger than Amazon Mechanical Turk that has far more work than you could ever tracked because when it comes to the accounting of this Workforce, it's effectively the equivalent of paper.
  • 25:26 - 25:30
    It's it's not considered labor Capital.
  • 25:30 - 25:37
    It's just an asset being bought and sold through a procurement Farm or through a procurement office.
  • 25:37 - 25:48
    So I say that coldly because importantly that's part of the legal framework in Arrangements that exist between businesses hiring outside of
  • 25:48 - 25:55
    Their companies to be able to have full time labor on tap through other firms that are the employer of record.
  • 25:56 - 26:04
    So it's a really important lineage to understand how it pipes into the back of several of these companies the other two companies.
  • 26:04 - 26:09
    We looked at organizations that we looked at lead genius is a social entrepreneurship.
  • 26:10 - 26:26
    It was a startup founded in Silicon Valley that generate sales leads and they have a global Workforce with a lot of recognition that trying to do the work that they want to have done in the United States is not something they could legally do without being classified as formal employers.
  • 26:27 - 26:37
    So they've moved a lot almost all of their Workforce outside of the United States and then a fourth organization that we studied Amara Amara on demand.
  • 26:37 - 26:44
    I'm fascinated by a more on demand and the one of the co-founders and organizers of it.
  • 26:44 - 26:47
    Dean Johnson is going to join us for a conversation later.
  • 26:47 - 26:48
    But what fascinates me
  • 26:48 - 27:06
    Now tomorrow is that it started out as a volunteer community again doing captioning and translation of video, which is technically a very hard problem to solve to be able to look at video in any robust way and interpret what are the actions that are happening in that video.
  • 27:06 - 27:22
    That's Way Beyond computation right now and this community of volunteers effectively became a magnet for companies that in organizations that wanted to be able to translate and caption their videos in other languages.
  • 27:22 - 27:26
    So company started approaching them saying can we just pay you to do this fast?
  • 27:26 - 27:39
    So we created a labor market that from it's very beginning was organized by the volunteer energy and the attention to the workers doing the work for this community.
  • 27:40 - 27:48
    And I think it's a wonderful example of what we could be doing differently in terms of organizing these worlds of work around workers themselves.
  • 27:49 - 28:05
    In looking at these four companies it became an opportunity to start from there and to put out surveys on each of those platforms to be able to reach workers themselves and ask them about themselves and at the end of those surveys to be able to ask them.
  • 28:05 - 28:17
    Would you be willing to meet in person for an interview and the map that you're looking at for work that is in theory work that can go anywhere that's available to anyone.
  • 28:17 - 28:19
    You should see some patterns.
  • 28:19 - 28:44
    There's something that maps onto the infrastructure of Outsourcing of places where there are not many job opportunities that are comparable to service work at retail stores or in other settings where the Pape the payment is about the same as what you might be able to get in these online markets so note the pattern because that should tell us there's something structuring again.
  • 28:44 - 28:46
    Who does this work where they do this work?
  • 28:46 - 28:48
    What are the other opportunities that it?
  • 28:48 - 28:52
    Forecloses or suggest or not available?
  • 28:53 - 28:58
    On top of those once we had enough people interested to do the interviews.
  • 28:58 - 29:04
    It just turned into old-fashioned anthropology in a lot of ways of go to where the people are meet.
  • 29:04 - 29:12
    People see who would be willing to allow us into their lives for a long enough period of time to be able to understand the ebb and flow of their engagement with this work.
  • 29:13 - 29:35
    So it becomes really important for example to be able to be in India and see what happens when the monsoon season hits and how people then whether no pun intended the work that they have to do which is effectively being online and having to be hyper-vigilant to pick up tasks that are then delayed by whatever might be getting in the way of their internet access, for example
  • 29:36 - 29:51
    The the layer and I you know hats off to Sid for figuring out the ways that he would be able to measure this world because I have to confess I wasn't I didn't really care that much about measuring I cared about if there's anybody experiencing this world that's enough for me.
  • 29:51 - 30:00
    And thankfully he helped me see there's a lot of value in being able to understand the distribution of this work the the
  • 30:00 - 30:06
    Patterns and so what I want to share with you is really the outcome of merging these two approaches.
  • 30:06 - 30:27
    What we found was in this is to me that one of the most striking findings in the book is that there's a real Pareto distribution of participation and when I say that like so many other power laws that are out there it turns out there's a concentrated few that are picking up and doing most of the work in these markets.
  • 30:27 - 30:35
    So a good depending on the platform a good ten to Fifteen twenty percent at most are picking up most of the tasks to be had.
  • 30:36 - 30:40
    And then there's this Core group of people that we call regulars that first group.
  • 30:40 - 30:52
    We call them always on and they literally are they've turned this into an income stream that maintains their livelihoods and they might have other income streams and often they were working on multiple platforms.
  • 30:52 - 30:56
    They've turned it into full-time work for themselves the second group.
  • 30:56 - 31:02
    There's about 30, you know, 30 percent at most but closer to 20 are what we call regulars.
  • 31:02 - 31:04
    They're stepping into this.
  • 31:04 - 31:08
    They've sunk the costs of figuring out how to make these platforms payoff.
  • 31:08 - 31:23
    They've learned what they need to learn and importantly they've connected with peers much like the always-on have they've connected with discussion forums other people who help them manage and figure out how to reduce their costs getting this work done.
  • 31:23 - 31:36
    They are the bench the Deep bench that is always able to step up into this labor market and pick up a task and do it and it is what allows anybody who's always on to step away and
  • 31:36 - 31:39
    Not have the entire Market just fall apart.
  • 31:40 - 31:58
    There's no way the argument we have here is there's no way to turn this into fully on time always on work to turn it into full-time work precisely works against what it is that people who have entered this world have said is important to them about entering this labor market.
  • 31:58 - 32:07
    I'm going to come to this in a moment, but lastly and most importantly there's a good 70% of people who walk into this we call them experimentalists.
  • 32:07 - 32:10
    Try it and they're like peace out don't want to do this.
  • 32:11 - 32:13
    They have a read of a range of reasons.
  • 32:13 - 32:15
    They decide they don't want to do it.
  • 32:15 - 32:24
    All of them are still providing value both for the companies that are able to claim that they have 500,000 workers on On Demand.
  • 32:25 - 32:45
    So think about any time you've used Lyft or Uber being able to see enough of those little cars that tells you okay, I'll bother that's the value all the experimentalists are bringing to this market and picking up one task or two tasks at is precisely being available that they are offering being willing and available.
  • 32:46 - 32:53
    That's the most valuable thing that they're doing and I think it raises this question of and yes, isn't that valuable?
  • 32:53 - 32:55
    Why isn't that considered valuable?
  • 32:56 - 33:02
    They're bringing an incredible amount of value certainly to the businesses.
  • 33:02 - 33:03
    They whether all of the costs.
  • 33:03 - 33:13
    These are most cases the people who could not figure out how to tap into a communication Network to make this manageable often felt isolated and alienated.
  • 33:14 - 33:25
    So collaboration is key in this environment the thing that allowed people to find their footing to be able to make enough money to make this worthwhile.
  • 33:25 - 33:32
    And again, the trade-off being being able to do it on particular terms was tapping into a network.
  • 33:32 - 33:39
    This is from an experiment that Sid Surrey and whenever co-authors manian developed to be able to identify.
  • 33:39 - 33:42
    What were the communities that people were engaged in discussion forums.
  • 33:42 - 33:47
    The different colors are the different discussion forums for this is just for Amazon Mechanical Turk.
  • 33:48 - 34:00
    So an incredibly robust Rich nuanced complicated environment of interaction and all of the small dots are this the solitary workers.
  • 34:01 - 34:18
    So the vast majority of people who haven't connected with somebody are there still providing value, but you've got this tight cluster and groups who are organizing around specific communities who again really scaffold with each other and make this manageable work.
  • 34:19 - 34:27
    I wanted to share this quote from one of the workers talking about how important it becomes to be able to connect with other people doing this work.
  • 34:28 - 34:36
    I think the Deep irony here is that the platform Builders assumed this is great work because you can do it alone and you don't have to interact with anybody.
  • 34:36 - 34:46
    You don't need any help what they hadn't anticipated perhaps because they didn't have enough anthropologists and sociologists in the room was that people might still be invested and having social connections.
  • 34:47 - 34:51
    The social connections are actually incredibly valuable to getting work done.
  • 34:51 - 35:06
    It's just immeasurably valuable and that's that's that's part of this environment is figuring out how to recognize the value of that connection the motivators that came up most often in this might sound trite.
  • 35:06 - 35:18
    It actually maps on entirely to the literature we have already about how people talk about their work what they value from work particularly when they decide to be self-employed or to try and freelance.
  • 35:18 - 35:18
    It's about
  • 35:18 - 35:25
    Trolling your time and I'd like everybody to stop using the word flexibility if you please because this isn't about flexibility.
  • 35:25 - 36:02
    It's about having other constraints on your time and needing to control your schedule often having to do with Family Care Elder Care Child Care other work responsibilities and other interests having interest in other education other other Hobbies other Joys and in most cases people making doing that calculus of how can I make this kind of work sustained me so that I can make my life run the way I want it to run that's aspirational to be sure but it's certainly part of what motivates people to keep at this
  • 36:03 - 36:18
    The second thing they're after is to be able to control what they work on and I'd imagine many people in this room share that they do just about anything to be able to Define what their project is rather than have somebody else tell them what to do and the third is controlling your work environment.
  • 36:18 - 36:37
    And if you're pushed to the margins controlling your work environment is not a nice to have so people with disabilities queer identifying people women who felt marginalized by the formal employment opportunities in their area all talked about this work being away of relief from those other constraints.
  • 36:39 - 36:48
    So when Carmela is talking about effectively being able to turn the this work because she can then pick up her computer and do it from anywhere.
  • 36:48 - 36:56
    I want to take very seriously and not dismiss her claiming that this is the kind of work that lets her live her ideal life.
  • 36:57 - 37:10
    It's how do we recognize take that at face value and still remained critical of a system that might still take advantage of her desire to do this work without somehow thinking that Camilla is the problem here.
  • 37:12 - 37:16
    So I want to move into thinking about where do we go from here?
  • 37:16 - 37:28
    What have we learned from the way these workers not only survive this work, but make it meaningful that we could then use to redefine this world of work because I'll continue to say this is early days.
  • 37:28 - 37:35
    We have an opportunity to really Design This purposefully with people at the center of our equation.
  • 37:35 - 37:38
    So I want to focus on two things if you
  • 37:38 - 37:41
    The book there's a the entire conclusion is just here's what we could do.
  • 37:41 - 37:55
    It's my bucket list that comes from the bucket list of the workers we engaged but the one I want to focus on for this conversation is to think about what it means to redefine the social safety net and job classification.
  • 37:56 - 38:02
    I'm just going to say it I would like to completely blow up employment classification as we know it.
  • 38:03 - 38:23
    I do not think the defining full-time work as a place where you get benefits and part-time work as a place where you have to fight to get a full-time job as an appropriate way of addressing this labor market and particularly if we consider that globally there are so few people in the world who have ever had access to full-time employment that provided any benefits.
  • 38:24 - 38:37
    So let's start organizing our classification and treatment of employment with all of the contingent work we've seen in this lineage and imagining that we will be those workers in the future and if that is the case
  • 38:38 - 38:42
    The best thing we could do is say there are some Basics here that we're building a Commons.
  • 38:42 - 38:50
    We have a labor Commons that every Company benefits from being able to draw from they can dip in and out of this pool.
  • 38:50 - 38:53
    So, how are we going to support that pool?
  • 38:53 - 38:54
    How are we going to make it sustainable?
  • 38:55 - 38:57
    So that the value proposition isn't here.
  • 38:58 - 39:12
    Let's just exhaust this pool and drain it because that's a tragedy we know of the tragedy of the commons apply the same logic imagine if health care for all isn't just a nice thing to do because a charitable it's like that makes business sense.
  • 39:12 - 39:14
    You need a healthy Workforce.
  • 39:14 - 39:20
    You need people to be able to step in and out of it to be able to make it sustainable continuing ed.
  • 39:20 - 39:27
    There wasn't a person that we interviewed who didn't talk about how often they were going to online resources to be able to continue exploring.
  • 39:27 - 39:34
    What were the kinds of materials they needed to follow up on and choose to read be able to do their neck next project.
  • 39:34 - 39:36
    We all benefit from being able to do that.
  • 39:37 - 39:38
    Everybody had a
  • 39:38 - 39:56
    Baseline of a liberal arts education the should be the big news is a liberal arts education and learning how to learn is the bass bass line for every worker in this market because it's creative work all of us had to learn how to learn as that Baseline and then build from there.
  • 39:56 - 39:58
    This is not specialized work.
  • 39:58 - 40:01
    This is using your brain all the time.
  • 40:02 - 40:13
    So the basic education that comes with critical thinking is key and then being able to make everything else available available becomes critical and then thinking about co-working spaces.
  • 40:13 - 40:32
    I just want to leave you with this Vision when I saw all of the home office setups of everybody we interviewed and at the end of the day thought of the health and safety administration and what our workplace around health and safety was set up to do it was to prevent Public Health crises.
  • 40:32 - 40:38
    We have a Public Health crisis and people setting up their home offices and not having resources to do it.
  • 40:38 - 40:39
    A way that's healthy.
  • 40:39 - 40:46
    So I know that sounds small, but every municipality could have co-working space that is thinking public health.
  • 40:47 - 40:49
    That's an intervention at the public health level.
  • 40:49 - 40:55
    Everybody should be able to get to a space where they can get relief for their back for their neck again sounds trite.
  • 40:55 - 40:58
    But if you're doing office knowledge work, it's critical.
  • 40:58 - 41:01
    And then lastly I want to dwell on this notion of a retainer.
  • 41:02 - 41:06
    So there's a lot of conversation about Universal basic income as a solution here.
  • 41:07 - 41:30
    What frustrates me most is the framing of that of that suggestion is that these are the poor souls who can't outrun the robots that doesn't really account for how much what we really need is to retain the cognitive and creative capacity of people to be available to businesses to each other for service work.
  • 41:30 - 41:37
    So people should be on a retainer for sure give everybody who's a working-age adults a retainer that says here's your base.
  • 41:38 - 41:42
    Line so that you don't have to think about paid leave or unemployment.
  • 41:42 - 42:05
    But that literally you have what sustains you to be able to step away when you need to step away have children take care of someone and at the same time know that you are going to have the financial means to get back into that comments right very different attitude than Universal basic income the idea that somehow the poor are going to rise against us.
  • 42:05 - 42:10
    And therefore we need to give them some some small amount of money completely misses.
  • 42:10 - 42:12
    Where is this economy headed?
  • 42:12 - 42:13
    It's a service economy.
  • 42:14 - 42:20
    It's an information service and Care economy where we're caring for each other and so to be able to do that.
  • 42:20 - 42:22
    We really have to imagine.
  • 42:22 - 42:28
    How would we give everybody the basic support financially to do that to be able to come and go in that work?
  • 42:29 - 42:39
    So the last thing we have to do is on us if you use any of these Services now is the time and today is a great day to do it.
  • 42:39 - 42:41
    There's a strike by Lyft and Uber drivers.
  • 42:42 - 42:59
    If you're a consumer of any ride hailing app be thinking about critically if you're somebody who consumes with care if you think about where you buy your clothes or buy your food, there's nothing small about that movement consumer advocacy and boycotting have been incredibly powerful.
  • 43:00 - 43:27
    They led to the Bangladesh Accord which was the beginning of holding companies accountable for the long supply chain involved in getting the shirts on our backs and in thinking about agriculture and places where knowing the supply chain became critical to making sure the quality of food, but the quality of people's work conditions growing that food was something everybody could know and then make your choices and that will never be enough.
  • 43:27 - 43:29
    This is not a land of
  • 43:29 - 43:35
    Businesses be kind to workers or let consumers be self-interested consumers.
  • 43:35 - 43:42
    It's about setting up the possibility for the right regulation and classification for employment of the future.
  • 43:43 - 43:48
    So the key takeaway other than a I always requires people.
  • 43:48 - 43:49
    So we're not getting rid of them.
  • 43:49 - 43:52
    Anytime soon is that this is a Commons.
  • 43:52 - 44:02
    This is a labor Commons that relies on people coming in and out and what will it take for us to be able to support and value labor?
  • 44:02 - 44:03
    No matter where it happens.
  • 44:03 - 44:11
    No matter how many hours somebody puts in because in fact the value of these markets it's aggregating up what we do.
  • 44:12 - 44:18
    It is literally the aggregation of everybody's input and from there what it's providing.
  • 44:18 - 44:22
    So think Uber--it's all of those drivers being available to you.
  • 44:22 - 44:25
    It's not just the one driver who took you to the airport.
  • 44:25 - 44:29
    So to be able to value all of the aggregation of that labor.
  • 44:29 - 44:38
    And say what will it take to say each of the people participating in that are equally important to our livelihoods to our lives.
  • 44:39 - 44:45
    And with that I want to thank my collaborators because there's nothing like a book about this kind of work
  • 44:45 - 44:51
    to make you hopefully always aware of how much-- everything that went into.
  • 44:51 - 44:56
    This book came from an
    incredibly rich team of people
  • 44:56 - 44:57
    bringing a range of expertise
  • 44:57 - 45:01
    from Greg Minton, who made
    the beautiful maps
  • 45:01 - 45:06
    to all the research assistants who were
    involved in doing the fact checking
  • 45:06 - 45:11
    to the key research assistants that we
    had in India to be able to maintain
  • 45:11 - 45:14
    the contacts with the people
    we had met through fieldwork
  • 45:14 - 45:19
    so there wasn't a person on this team
    who didn't do something integral that
  • 45:19 - 45:23
    if they weren't here I don't know that this
    book would be here either, so with that,
  • 45:23 - 45:24
    thank you
  • 45:24 - 45:32
    [Applause]
  • 45:38 - 45:40
    Thank you Mary that was amazing
  • 45:40 - 45:44
    I'd like to invite Dean Jansen to
    join us at the front of the room
  • 45:44 - 45:45
    Dean is the executive director
  • 45:45 - 45:50
    and chief executive officer of the
    Participatory Culture Foundation
  • 45:50 - 45:53
    The parent organization of
    Amara On Demand
  • 45:53 - 45:57
    co-leading PCF with Aleli Alcala
  • 46:00 - 46:01
    Amara is featured in Ghost Work
  • 46:02 - 46:04
    and we're thrilled to welcome
    him into the conversation.
  • 46:05 - 46:07
    so as moderator I'm
  • 46:07 - 46:09
    gonna take the host privileges of asking
  • 46:09 - 46:12
    a few questions of Dean and then Mary
  • 46:12 - 46:15
    and then I'll open it up to
    the floor for a discussion
  • 46:15 - 46:19
    Dean maybe I'll start with you
    if you don't mind OK?
  • 46:19 - 46:22
    If you could tell us
    a little bit about what's
  • 46:22 - 47:11
    changed for Amara
    since the book was
  • 47:11 - 47:14
    sure yeah this was a great question
  • 47:14 - 47:50
    and one that you know
  • 47:50 - 47:52
    thinking about what has changed
  • 47:52 - 47:55
    as an organization
    we've gotten bigger
  • 47:56 - 47:57
    When we started and when
    kind of AOD was first
  • 47:58 - 47:59
    starting I think we were a staff of
    about 9 to 12 people
  • 48:03 - 48:04
    the initial group of folks that
    were doing the translation
  • 48:13 - 48:13
    around 200 and today,
    the staff is closer to 30
  • 48:13 - 48:13
    there are thousands of different
    folks who have signed on
  • 48:13 - 48:13
    and her are working with with AOD at this point
  • 48:13 - 48:14
    and then as far as
    just them the marketplace
  • 48:14 - 48:14
    that's something that you know it
  • 48:14 - 48:14
    when we began there was obviously
    a robust translation marketplace
  • 48:14 - 48:15
    but the audiovisual subtitling and
    dubbing side of things
  • 48:16 - 48:17
    was still pretty nascent and
  • 48:17 - 48:18
    that's really starting to pick up
    more and more today
  • 48:18 - 48:19
    so those are two areas that
    have changed significantly.
  • 48:23 - 48:24
    Mm-hmm.
  • 48:25 - 48:26
    one thing that Mary
    talks about in the book
  • 48:26 - 48:27
    is this idea of the double bottom line
  • 48:27 - 48:28
    and she talks about that
    specifically when it comes to Amara
  • 48:28 - 48:29
    which tries to create fair and
    collaborative labor practices for people
  • 48:29 - 48:30
    who are doing translation
    and subtitle work
  • 48:30 - 48:31
    and so what I wanted to ask you
    Dean is
  • 48:31 - 48:31
    how do you respond
    to people who say that
  • 48:31 - 48:31
    the double bottom
    line is a nice idea
  • 48:31 - 48:32
    but it's really not practical...
  • 48:32 - 48:32
    that in fact a business like yours can't
    sustain itself with that model in place
  • 48:32 - 48:32
    yeah that's that's a
    great question and one that I think
  • 48:32 - 48:32
    might be answered well with another
    question, which is,
  • 48:32 - 48:32
    what does it mean to be sustainable
  • 48:32 - 48:33
    from organization to
    organization ?
  • 48:33 - 48:33
    Again, in the in the space
    that we're in,
  • 48:33 - 48:33
    in translation,
    we saw early on
  • 48:33 - 48:33
    these kind of more established
    players that had very high margins.
  • 48:33 - 48:39
    And so again, just a question of
    how do you define sustainability?
  • 48:39 - 48:39
    Sustainability for whom and for what?
  • 48:39 - 48:40
    I think if you're asking more maybe about
    the lower margin end of things
  • 48:40 - 48:41
    we're talking about some companies
    that are the biggest and
  • 48:41 - 48:55
    most powerful and profitable
    in the history of the planet.
  • 48:56 - 49:04
    And so I think it's a question less
    – in our eyes at least –
  • 49:05 - 49:06
    of "is it sustainable?" and more
    a question "make it sustainable?"
  • 49:07 - 49:12
    I'll just name MTurk as
    one of those companies
  • 49:12 - 49:13
    since Dean was too
    polite to say it.
  • 49:13 - 49:14
    Another question I
    wanted to ask about is
  • 49:14 - 49:20
    really playing off of what Mary was
    talking about at the end of her talk.
  • 49:21 - 49:22
    We can think about consumer advocacy,
  • 49:22 - 49:36
    and we can think also about regulation.
    In your mind from where you sit
  • 49:36 - 49:37
    what are the possibilities and limits on
    what a business enterprise can do
  • 49:37 - 49:37
    to value and protect micro labor?
  • 49:37 - 49:40
    Hmm another great question.
  • 49:42 - 49:44
    I think, well, let's see.
  • 49:44 - 50:11
    So as far as the limitations go,
    I think there are a lot of things
  • 50:11 - 50:12
    that organizations can
    and should be doing.
  • 50:12 - 50:12
    We were kind of joking about
    – not joking but –
  • 50:12 - 50:12
    a couple of weeks ago, we were
    talking about how low the bar is,
  • 50:12 - 50:12
    where the bar is to recognize that
    these are human beings doing this work.
  • 50:12 - 50:12
    Early early on when
    speaking with Mary,
  • 50:12 - 50:12
    one of the things that really struck me
    was her describing conversations
  • 50:12 - 50:20
    she would have, as she
    mentioned, with engineers
  • 50:20 - 50:21
    who really didn't see where all these
    layers existed with human labor in them.
  • 50:21 - 50:38
    So to me and many ways the work that
    Mary and Sid and all the people
  • 50:39 - 50:39
    that have made this this book possible
    all the people that Mary spoke with
  • 50:39 - 50:40
    shining a light on it and making it
    visible as the first step in figuring out.
  • 50:40 - 50:41
    I think I've just
    skipped your question,
  • 50:41 - 50:48
    but in terms of zooming
    out and looking at societally
  • 50:48 - 50:49
    what can we do and how can we
    accomplish some of these things?
  • 50:49 - 51:03
    Just that first step
    of having some recognition
  • 51:03 - 51:04
    of who and where people
    are is really important
  • 51:04 - 51:04
    but obviously on an individual,
    organizational level
  • 51:04 - 51:05
    there is a ton that can
    be be done
  • 51:05 - 51:05
    and you give us a few,
    just a few examples from
  • 51:05 - 51:12
    you're sure yeah,
    so let's see in terms of
  • 51:14 - 51:26
    providing people space
    to communicate
  • 51:26 - 51:27
    and work with
    one another
  • 51:27 - 51:28
    one of the things that we
    found in translation was
  • 51:28 - 51:29
    there's this kind of stereotype
    of the Lonely translator
  • 51:29 - 51:29
    someone who is working
    in an isolated sort of and in many ways.
  • 51:29 - 51:30
    That's again, I'm not a translator.
  • 51:30 - 51:32
    So this is just my learning
    and understanding of it but
  • 51:32 - 51:44
    historically people do it has
    been a more isolated kind of work.
  • 51:44 - 51:59
    So for us bringing the
    collaborative side of the tools
  • 51:59 - 52:00
    and platform that we're
    providing and building
  • 52:00 - 52:00
    to continue listening to what
    sorts of things people need
  • 52:00 - 52:00
    to do their work better and
    to collaborate more effectively.
  • 52:00 - 52:04
    That's been something that's
    been the bread and butter
  • 52:04 - 52:05
    of some of what we've been doing.
  • 52:06 - 52:08
    Thank you being cognizant of time.
  • 52:08 - 52:12
    I will just ask Mary
    a few short questions
  • 52:12 - 52:12
    and then open
    it up to the floor.
  • 52:12 - 52:17
    So Mary, one of the key
    concepts in your book.
  • 52:17 - 52:20
    Is this paradox of
    "automation's last mile."
  • 52:20 - 52:34
    Could you pull that out for us
    a little bit and maybe relate it
  • 52:34 - 52:35
    to the lineage of contract contingent
    labor that you put up for us
  • 52:35 - 52:36
    which in your book actually
    very interestingly starts with
  • 52:36 - 52:36
    the experience of slavery
    in the United States.
  • 52:36 - 52:50
    Yeah, I mean, to think about this
    paradox, in many ways,
  • 52:50 - 52:50
    is to grapple with what has been the
    use of the Lion's Share of manual labor.
  • 52:50 - 52:52
    For example, and in early days,
  • 52:52 - 52:52
    if we think about what
    defines modernity,
  • 52:52 - 52:52
    what defines our modern era,
  • 52:52 - 52:58
    It's imagining we're going to
    get our hands out of the soil
  • 52:58 - 52:59
    and be able to put
    our minds to work.
  • 52:59 - 53:07
    It's an erudite notion of what
    does advancement look like,
  • 53:07 - 53:08
    what is progress and
  • 53:08 - 53:08
    progress driven by
    technology?
  • 53:08 - 53:14
    And so the paradox is that,
  • 53:14 - 53:15
    as we strive to pull
    ourselves out of manual labor,
  • 53:15 - 53:25
    I believe we start recognizing
    that it actually takes quite a bit
  • 53:25 - 53:26
    of creativity and complexity
    to be able to do any enterprise...
  • 53:26 - 53:27
    to be able to do
    anything productive.
  • 53:27 - 53:33
    So starting the lineage
    with slavery in the book
  • 53:33 - 53:34
    is to say, that was really our
    first labor law in the United States
  • 53:34 - 53:35
    it found who could be owned,
    who could be used
  • 53:35 - 54:09
    in a way that only treated
    their bodies as valuable
  • 54:09 - 54:10
    and didn't imagine that
    any kind of work that we do
  • 54:10 - 54:11
    involves a human capacity.
    Not to be too humanist here
  • 54:11 - 54:12
    but a human capacity to be
    able to bring creativity,
  • 54:12 - 54:13
    to be able to bring responding to
    spontaneity to whatever we're doing
  • 54:13 - 54:13
    so that moment of delaying recognizing
    the value of the real deep integration
  • 54:13 - 54:14
    and of creativity with
    everything that we do
  • 54:14 - 54:14
    I believe that is this moment.
  • 54:14 - 54:17
    It's this reckoning
    with a paradox.
  • 54:17 - 54:22
    We keep introducing and
    thinking we're just going to
  • 54:22 - 54:23
    we're going to get the thinking
    and the talking out of it.
  • 54:23 - 54:29
    It's just going to be
    something we can automate
  • 54:29 - 54:32
    and then we're just left with
    things that are really hard
  • 54:32 - 54:34
    the professional class will be able
    to do the really hard work
  • 54:34 - 54:35
    it throws to the wind the idea
    that taking care of a parent...
  • 54:35 - 54:40
    if you've ever taken care
    of an elderly parent...
  • 54:40 - 54:44
    that takes a lot of thinking,
    a lot of creativity.
  • 54:44 - 54:55
    There's a beautiful book that sets up
    the discussion of the paradox
  • 54:55 - 55:00
    and it's by Levy and Murnane, they're
    computer scientists and economists
  • 55:00 - 55:03
    who talk about the
    new division of labor
  • 55:03 - 55:04
    and they were trying to analyze
  • 55:04 - 55:05
    what is it that a computational
    process can do that a human can't
  • 55:05 - 55:06
    and what is it that a human
    can do that is really distinct?
  • 55:06 - 55:07
    What is that division?
  • 55:07 - 55:13
    And so we're trying to to theorize
    why it is as we strive to automate things.
  • 55:13 - 55:21
    We keep discovering anew
    that the capacity to think creatively
  • 55:21 - 55:25
    is in the thick of it and that
    as we keep pushing to automate
  • 55:25 - 55:29
    and reaching for having
    something else do things for us.
  • 55:29 - 55:32
    We'll keep discovering those bits
    that are really the heart and soul
  • 55:32 - 55:35
    of what humans do...
  • 55:35 - 55:39
    ... which is sense each other's needs,
    anticipate, and––I joke often––
  • 55:39 - 55:43
    be able to apologize when we get it
    wrong. Computation can't do that.
  • 55:46 - 55:47
    I'm going to just ask one last question
  • 55:47 - 55:50
    I know there's a lot of people waiting
  • 55:50 - 55:53
    One thing I wanted to underline
    which I loved in your presentation
  • 55:53 - 55:58
    and in the book
  • 55:58 - 56:00
    is the way that you use the
    task of Mechanical Turk itself
  • 56:00 - 56:02
    to set tasks to get
    research for the book.
  • 56:02 - 56:04
    I think this is a
    brilliant methodology.
  • 56:04 - 56:13
    So I just want to put
    a big fat line under that
  • 56:13 - 56:17
    and then and then based
    on that I wanted to ask you
  • 56:17 - 56:19
    if you could pull out for us
    the significance for you
  • 56:19 - 56:21
    in doing transnational
    comparative work,
  • 56:21 - 56:22
    especially as it relates to
    identifying the challenges
  • 56:22 - 56:23
    and opportunities in
    labor organizing.
  • 56:24 - 56:34
    To be clear, starting with India
    and the United States
  • 56:34 - 56:40
    in some ways was following
    the labor markets that
  • 56:40 - 56:43
    Amazon Mechanical Turk had created
    by paying in both cash and rupees;
  • 56:43 - 56:44
    they had built this labor
    market by design
  • 56:44 - 56:45
    without probably much thought
    about the kinds of work forces
  • 56:45 - 56:45
    it would create.
  • 56:45 - 57:01
    In drawing that comparison
    it meant we were constantly
  • 57:01 - 57:09
    able to pull out the places
    where connection broke down
  • 57:09 - 57:13
    the kinds of nationalisms that would
    spark around different work groups
  • 57:13 - 57:15
    For example, the transnational
    comparison gives us a chance to see
  • 57:15 - 57:16
    where two people throw
    boundaries back up
  • 57:16 - 57:16
    I thought it was most
    striking in India
  • 57:16 - 57:17
    how quickly a kind of planned in the end way of orienting to work across all the platforms also came to the Forefront.
  • 57:17 - 57:23
    So the number of women and men who would start referring to each other as sister and brother to navigate the gender politics of
  • 57:23 - 57:59
    King and settings with somebody of a different gender the the amazing use of English as a way of navigating linguistic boundaries and what that can mean the meshing and sidestepping of religious and cast differences those complexities could come out and in some ways being able to see the complexities around class and race that played out in the United States in different ways really came to the fore with that comparison the technique of putting the surveys online.
  • 57:59 - 58:09
    I think in many ways. I'm really dedicated to us always imagining that anthropology sociology of engaging people's lives means getting in their lives.
  • 58:10 - 58:20
    I have a I have a difficult time feeling like stopping at a discussion forum and letting that stand in for people's experiences can answer the questions.
  • 58:20 - 58:23
    I'm interested in I think it has everything to do with the question you're asking
  • 58:23 - 58:28
    For the most part the question we were asking was what does the rest of your life look like?
  • 58:28 - 58:41
    And so there was really no way to ask that without moving from those platforms into the their living rooms into the cafes they circulated in and that that became the the methodology.
  • 58:42 - 58:47
    Thank you who would like to ask a question Dana?
  • 58:51 - 58:53
    This is Mary's fault. She set me up.
  • 58:53 - 59:23
    She was like you have to ask a question was like I was all right ready so much of what you're grappling with is a dynamic that we've seen iterate over and over again throughout history and you've pointed to them by talking about the enslaved people and as a form of contract with in labor markets where different versions of capitalism have evolved in response to these different governing structures, and we keep seeing a regulatory move and capitalism evolved.
  • 59:24 - 59:50
    We are now at a late stage capitalism structure where we're not only seeing the efficiencies that are produced by capitalism as as an operating system, but reinforced as you point out by Technical Systems, but we're doing it in a environment to sort of Riff Off of Cerritos point where we're not actually dealing with bounded nation state structures, and I've noticed you're not screaming to tear down late-stage capitalism and maybe a
  • 59:50 - 60:00
    I know you but part of it is how do we actually think about the boundary work of nation state structures and the evolutions of late-stage capitalism.
  • 60:00 - 60:04
    Something that can actually grapple with this so that it doesn't just keep slipping in evolving.
  • 60:07 - 60:14
    Two things popped in my head the last week or so one is thinking multinationals have figured this out in some ways.
  • 60:14 - 60:35
    They know how to cash flow and so to take at face value that multinationals generating quite a bit of wealth internationally and circulating it internationally can't figure out how to circulate and distribute the value seems hell.
  • 60:35 - 60:41
    No. No, they should not govern know it is the say there's clearly a way to exchange money.
  • 60:41 - 60:49
    Globally. I'll just I'll say it flatly like that and that means in terms of governing the revenue generated the redistribution of it.
  • 60:49 - 60:50
    That is a global conversation.
  • 60:51 - 60:57
    We know that multinationals are regulated in a specific company or in specific countries start there.
  • 60:58 - 61:05
    So make the United States make the EU the Battleground for saying new classification.
  • 61:06 - 61:08
    Everybody gets Basics go from there.
  • 61:09 - 61:29
    I think it's for me very frustrating and realize how many large companies and we're pretty much in a monopolistic, you know World here how many of the large companies that have merged and acquired each other have been able to stay on the sidelines of a conversation around Universal Health Care in the United States that makes no sense.
  • 61:29 - 61:42
    And it should be on everybody's mind to be advocating to these companies and it can't be about just advocacy but certainly making the case why are you on the sidelines you benefit from being able to see this healthy Workforce?
  • 61:42 - 61:57
    And in fact one way to think about this as Supply chains and good work codes for Enterprises to say there should be a call for regulation that if you are hiring a vendor you should make sure that vendor provides.
  • 61:57 - 62:02
    Of care just as you are required or at least benefiting from providing Health Care.
  • 62:02 - 62:44
    There's a business case there can it happen without regulation know we all need to say this is ridiculous that we don't provide basic Healthcare and I would I would just keep making that case but the second thing that came to mind is that it really is on all of us to stop letting liberal and neoliberal economics to find the value of Labor to call Labor and Marketplace and not see the moralism within that like, I I'm just going to spell this out just plainly we have allowed people in Elite positions both government and private Enterprise to say that our labor is the same kind of capital as a set of Records or a car.
  • 62:45 - 62:48
    We're not durable goods labor means something else.
  • 62:48 - 62:53
    So let's stop assuming that the marketplace sets the value of our labor.
  • 62:54 - 62:56
    My salary is not coming from some magically.
  • 62:57 - 63:00
    Earned value of me on a Marketplace.
  • 63:00 - 63:01
    It's coming from Power.
  • 63:02 - 63:11
    I have a particular kind of power and privilege to command a price and that is that as a set of irrational power moves.
  • 63:11 - 63:14
    It's not the logic of the market playing out.
  • 63:14 - 63:17
    So I just I really want us all to start questioning.
  • 63:17 - 63:19
    There should be no turn to the market.
  • 63:19 - 63:22
    What is the market paying for this task?
  • 63:23 - 63:32
    It should be Baseline labor divert deserves support because we need healthy workers to move the world forward.
  • 63:32 - 63:38
    I'm a pragmatist. That's why I'm not that interested in blowing up capitalism writ large.
  • 63:38 - 63:47
    I don't know you could probably get me there Dana but there are alternative versions of marketplaces like Cooperative marketplaces marketplaces.
  • 63:47 - 63:57
    We haven't even imagined yet because we haven't let ourselves think that the market can't set the value of us nor should it and that's why that lineage
  • 63:57 - 64:04
    It was slavery. Why would we have ever imagined that the market or a market should Define the value of humans?
  • 64:04 - 64:13
    Let's stop that and and and set a Baseline and then go from there perks sure you get perks if you're an extra good worker.
  • 64:15 - 64:16
    Yeah.
  • 64:18 - 64:21
    Last question one more of those.
  • 64:23 - 64:25
    I think I might have a simple question.
  • 64:25 - 64:43
    But I'm I'm sort of thinking as you've been talking about people's power and how we might choose a baseline that might be a retainer or some sort of work for all do you think that that might be really difficult as more and more people are doing labor unwittingly.
  • 64:43 - 64:52
    Like for example people like Google uses verification to make sure you're not a robot and we're like great.
  • 64:52 - 64:59
    That's great. I want to improve on human but you're actually doing work for for that you're doing that task and you don't know that you're doing it.
  • 64:59 - 65:08
    So do you think that like Hulu or subscription services are going to start making people do work in order to watch a video and because they want to watch it.
  • 65:08 - 65:11
    They're going to do work like in these microtransactions.
  • 65:11 - 65:18
    Do you see that maybe killing the labor force to because all the works going to be spread out to everyone that are very willing to like
  • 65:18 - 65:20
    Watch their content to do it.
  • 65:22 - 65:35
    Really optimistic reasons that I think that won't happen one is that when because when people become more aware that when you check that, I'm not a robot that you're actually doing work for a company that makes a lot of money and doesn't actually make your service better.
  • 65:35 - 65:54
    We should all be saying he'll know and and it's not that we're going to walk away from it because we're all enjoying those Services is that we're all going to become publicly aware that that you are having your time taken you are having your time taken without your permission.
  • 65:54 - 65:56
    That should never be. Okay, right.
  • 65:56 - 66:04
    So one one answer to your question is we absolutely have to have more public awareness of the sheer robbery of our time.
  • 66:04 - 66:06
    That's that's happening.
  • 66:06 - 66:08
    We make pains the second response to your question.
  • 66:08 - 66:12
    We make pains in the book to distinguish between paid labor and unpaid labor.
  • 66:12 - 66:18
    And the reason to focus on paid labor is precisely say that is a different relationship.
  • 66:18 - 66:22
    It is a social or it's not a charity when you get a job from somebody.
  • 66:22 - 66:24
    You are providing value.
  • 66:24 - 66:34
    It's a business transaction be cold about it so that we can start living our lives and stop having work Define who we are right?
  • 66:34 - 66:41
    Like this is Kathy weeks is one of my favorites on this like, why are we not fighting for the end of work rather than for a 40-hour workweek?
  • 66:41 - 66:43
    Hell no, right.
  • 66:43 - 66:53
    So let's go there and I think the beginning of it is with an awareness of when we are giving away our and actually I think that's the wrong framing when our time is being taken from us.
  • 66:54 - 67:03
    Let's stop saying we're giving it away because our time is being taken from us without her permission and we have to we have to say that's not okay.
  • 67:04 - 67:05
    That's the first part.
  • 67:07 - 67:09
    Okay last question.
  • 67:15 - 67:17
    Thank you guys. This is just great.
  • 67:17 - 67:46
    So I wanted to ask about I'm looking at the cover of the book and intensely wondering who is this person and and it but it brings back the question the topic of the visibility of people doing this work and their their relative invisibility and I wondered if you could talk about that and talk about why or how that visibility might be important to the bigger equation of change that you imagine.
  • 67:50 - 68:03
    There's not a day that goes by that somebody doesn't tell me about their lift or their Uber driver when I talk about this book and so there's clearly something poignant and and pressing for people when the person that they're worried about.
  • 68:03 - 68:04
    Is it right in front of them?
  • 68:04 - 68:12
    I mean and that's a good thing and I think in some ways let's tactically deploy that if you're an organizer, I do think we care.
  • 68:12 - 68:14
    I mean, I know I'm a roll.
  • 68:14 - 68:18
    Was Optimist I think when we see someone in pain we want to relieve it.
  • 68:19 - 68:20
    Usually we save that for people.
  • 68:20 - 68:23
    We love the most but our
    impulse is there.
  • 68:23 - 68:32
    And so I want to believe that,
    by raising visibility...
  • 68:32 - 68:36
    by seeing the people
    who do this work,
  • 68:36 - 68:38
    it changes what we want for them.
  • 68:38 - 68:39
    You could just be crass and cynical
    and say this could be you
  • 68:39 - 68:40
    so don't you want to change
    your circumstances?
  • 68:40 - 68:40
    There will absolutely
    be someone in your life.
  • 68:41 - 68:43
    Do you want to see
    their life better?
  • 68:43 - 68:44
    So I think it is for me.
  • 68:44 - 68:49
    It's funny you raised that, I was
    uncomfortable with that cover.
  • 68:50 - 68:51
    I found it sensationalistic.
  • 68:51 - 68:52
    I'll just say this. I love my process.
  • 68:52 - 68:54
    I don't know if they're here right now.
  • 68:54 - 68:56
    Clearly what they wanted was
    to get this book in people's hands
  • 68:56 - 69:00
    and they wanted something
    that people might reach for
  • 69:01 - 69:03
    And it might have that effect.
  • 69:03 - 69:06
    I hope it does, that was
    their reason for me.
  • 69:06 - 69:11
    I'm now on this mission to find out
    who is he and I'll get back to you.
  • 69:11 - 69:14
    I'll post something I'll find out who he is.
  • 69:16 - 69:19
    Yeah, and that is actually
    a really complicated question
  • 69:19 - 69:22
    because at the end
    of the day, for the people
  • 69:22 - 69:25
    I interviewed, they're folks doing
    their jobs, living their lives.
  • 69:25 - 69:28
    They're not particularly
    interested. For the folks I've asked
  • 69:28 - 69:32
    "Do you want to be on
    any of the TV shows...
  • 69:32 - 69:43
    ... I might be on" and the
    first thing they say is no
  • 69:43 - 69:43
    and the second thing they'll
    say is "I don't have time"
  • 69:43 - 69:44
    and the third thing
    they'll say is...
  • 69:44 - 69:44
    "I don't want to be seen as one
    of those people who's a victim."
  • 69:44 - 69:58
    So what I don't really appreciate
  • 69:58 - 69:59
    as most of the coverage out there
    frames the people who do this work
  • 69:59 - 70:00
    as victims or dupes
    or unaware of their circumstances .
  • 70:00 - 70:01
    The people doing this work are
    painfully aware of their circumstances.
  • 70:01 - 70:03
    And we would do best
    by listening to them.
  • 70:04 - 70:06
    I think we could just listen
    rather than need to see them.
  • 70:06 - 70:10
    This isn't really about
    everybody who's had a hand
  • 70:10 - 70:11
    in making your
    social media palatable.
  • 70:11 - 70:16
    It's about knowing there are
    people who are helping you
  • 70:16 - 70:17
    and they are working.
    So how do we support them?
  • 70:20 - 70:23
    Thank you very much
    buy the book so over here.
  • 70:24 - 70:25
    Thank you Dean.
Titolo:
Databite No. 119: Mary L. Gray
Descrizione:

more » « less
Video Language:
English
Team:
Volunteer
Duration:
01:10:35

English subtitles

Incompleto

Revisioni Compare revisions