A Diamond in the Economic Rough
Two weeks ago, the American Economic Association held its annual meeting in Philadelphia. Spread over three days and two different hotels, this conference included over 500 sessions.
Buried in the program, during the morning on the last day, was a grab bag paper session titled Radically Rethinking Economic Policy. The final paper discussed in this session should command our attention, because its coauthors include, in addition to four well-respected economics researchers, someone who I’ve long promoted as one of the most brilliant and outrageous thinkers pondering the digital world: Jaron Lanier.
The paper, which is titled “Should We Treat Data as Labor? Moving Beyond ‘Free’,” translates the basic premises of Lanier’s 2013 book, Who Owns the Future?, into more precise economic terminology.
This short paper contains many points worthy of longer discussion. I’ll focus here on its main conceit.
The authors note that a core resource of the digital economy is the data produced by users of services like Facebook and Google, which can then be used to train machine learning algorithms to do valuable things like precisely targeting advertisements or more accurately processing natural language.
The current market treats data as capital: the “natural exhaust from consumption to be collected by firms” for use in training their AI-driven golden gooses.
Lanier and company suggest an alternative: data as labor. Put simply, if a major platform monopoly wants your data to help build a multi-billion dollar empire, they must pay you for it. Offering a free service in return is not enough.
The idea of treating data as labor, which might require massive intervention in the form of user unions and government regulation, is an example of a radical market, an emerging concept in economic theory in which new markets are constructed specifically to act as countervailing forces to existing markets that are causing trouble.
To Lanier et. al., the data as labor radical market could defend against our worse fears about AI stealing our jobs by more equitably distributing the spoils of this next technological boom — supporting a middle class of “data workers” who find dignified employment in providing the quality data the boom will require.
To be clear, I’m not fully convinced. As Lanier himself admits, more formal analysis is needed to explore whether the vague notion of a dignified data worker class is actually feasible (regardless of the difficulties inherent in enforcing such a system). But I love the discussion.
The attention economy is not a minor bubble or teen fad, it’s a major force in our society that deserves the type of serious analysis that thinkers like Lanier and his economist coauthors are demonstrating. I hope this trend of careful but radical thinking continues.
#####
For a nice introduction to Lanier, I suggest Ezra Klein’s recent podcast interview, as well as Lanier’s epochal 2010 manifesto, You Are Not a Gadget. He also just published an acclaimed memoir titled Dawn of the New Everything.
(Photo by Luca Vanzella.)
This concept of data as labor is one that I also find to be an extremely interesting one to explore. Perhaps it ends up being unfeasible, but given that the current model seems highly likely to lead to even greater concentrations of wealth and power in the hands of a minority, new models need to be explored.
I’m a recent reader of this blog series and, admittedly, an instant fan having also just read Deep Work over the holiday–twice. The book couldn’t have come at a better time, as it provided just the words I was lacking to express (and explain) my growing discomfort with social media. I am particularly drawn to the emerging discussions around the social, cultural, and economic impact of the internet, consumer data, and social media (I work in the data and analytics space). Anyway, thank you for another thought-provoking commentary–and for the introduction to Jaron Lanier’s world of thought via the Ezra Klein podcast.
Don’t you think this would be totally counter-productive? In all of your writing you promote the benefits of a “deep life,” both in terms of psychological health and productivity, yet the system you describe here would actually provide an economic incentive for scattered attention among the masses. Remember that the vast majority of the data collected from consumers is providing lots of economic value without any societal value, much like cigarettes. To give the masses economic remuneration for their data would be like giving smokers discounts on cigarettes to reallocate Philip Morris’s profits. The problem is not who is getting the profits from causing cancer, but the causing of cancer! Any productive policy to deal with digital distraction should be working towards facilitating focus, depth, and psychological well being.
This is, I think, the blindspot in Lanier’s techno-philosophy: he always starts from the assumption that people have to use these technologies, so the best we can do is engineer them to be more rewarding.
Love the cigarettes analogy!
Lanier et. al. are saying that in the current one-buyer system it’s very difficult for users to ‘monetize’ their data except for bartering it for ‘free’ services (and hence there is little incentive for users to take ownership of their data). Dominating statistical ‘big data’ approach to AI/ML makes individual’s (small) data kinda useless outside of the ‘siren servers’, while the nature of the data (related to consumption rather than production) prevents AI from contributing to productivity growth.
I’m working on a ‘radical’ approach where user’s productivity data (provided by digital productivity tools – calendar, to-do list, time tracking) processed by non-statistical ML methods would offer individualized productivity advice. I see this as the flip side of ‘attention economy’ coin – our attention is focused into production and creation, rather than dispersed into consumption.
” In practice, paying customers would amount (from the company’s perspective) to a tax on selling aggregate data–and drive companies away from it, at least to some extent.”
It would also amount to an incentive to use the services more! It is entirely possible that the uptick in use would more than offset the cost of the payments, and actually increase profits. I also want to quibble with your use of the word “customer.” The customers are the companies buying ads(or data, etc), not the users. This is a large part of the problem.
I think there’s a possible middle ground. The YouTube channel “The King of Random” recently had a discussion about exactly this sort of thing. To monetize social media effectively, you need a multi-tiered approach. You need one group generating content, while another group orchestrates the social media side (picks which content to present, determines how to present it, etc). If you approach it from the perspective of a business, it’s possible to make a living off it, at least to a certain extent.
As for individuals, I think the more significant effect would be to create an economic disincentive for companies to sell your data in the first place. In practice, paying customers would amount (from the company’s perspective) to a tax on selling aggregate data–and drive companies away from it, at least to some extent.
Thanks for your comment. I tried looking on the King of Random page however isn’t obvious what the video was. Any chance you can remember or see what it was? Thanks.
Great idea. I remember thinking exactly this some years ago. What if FaceBook (or similar site) passed on some of that advertising revenue to the people that use that site. And then charge people to use FaceBook. The sadness is that people would then stop appreciating the social aspect of FaceBook and get ansty about not getting enough money from FaceBook for letting their data get to people and serve up all those adverts. And get up-in-arms if one month, they were on holiday not using FaceBook, and they had to pay that month, just because they *didn’t* use it.
This one’s going to be a difficult one to apply without making people think of their social communication as a commodity – not necessarily correlation to create.
This problem might be solved more elegantly by decentralizing applications and people be both users and stakeholders in the platforms. See Steemit, dtube, EOS etc.
Cal, check out the latest Art Of Manliness podcast wth Franklin Foer. His research and new book are right up your alley.
Have you read Tim Wu’s The Attention Merchants? It is the best book I have read so far about the economics of attention, comparing and contrasting attention-based models today and throughout modern history.
I think eventually the only way to live with AI – isn’t to beg it (and its owners) for scraps (by getting paid for using computers and consuming information) – it is to become part AI. Meaning, we will be part cyborg……eventually merge our consciousness with the tech.
That is how we will stay “useful” after AI has become powerful. That is how the middle class will return.
Sorry but no. Transhumanism is not the answer.
I’m a bit concerned about this. In a lot of cases, what people want from platforms such as Facebook is a way to share data–to share photos of kids/grandkids, to share stories, to connect with colleagues. Requiring Facebook to pay me for uploading photos of my kids would have a chilling effect on my capacity to upload pictures of my kids. This removes a real value from my life, and that of those who look at those photos (ie, my family).
A better model would be for Facebook et al. to stop viewing their users as products to be sold.
Cal – Apple’s CEO Tim Cook discussed technology overuse in a new interview, here’s a link of interest:
https://www.macrumors.com/2018/01/19/apple-ceo-tim-cook-technology-overuse/
I like the Idea of the “Data Labor”, but I don’t think it can adequately cover the large scale at which data recovering is being received. A question i have would be, wouldn’t this upset the whole system as it’s set up? They have the data for our phones, computers, etc. They offer it and and buy it. We accept that there is a transfer of information, which the majority are then ok with them using to improve our services. It would seem the opposite for them to pay for us to use their service. Where would the profit be in that? the only reason they can improve and develop their service is because they are receiving lots of money to run it. I can’t imagine them paying for people to use their service unless it’s a very specific way to use it for which they need people.
this idea might backfire as you have the opposite of the desired effect.anyway thanks for the sharing
If data were to be treated as labor, I wonder if there could be any implications regarding instituting (a small) universal basic income.
The future is a completely digital economy, updating is the main task that everyone must learn.
I read the same essay on Medium.com, it is really worth the attention, it is written by an intelligent person and I am sure that I will remain in my memory for a long time. The world will constantly change, whether we like it or not, you need to accept everything as it is and try to make the world better, so we constantly have to learn something new. I can also agree that the digital economy is our future.
I think that the digital economy is our future that is getting closer and closer to us, but at the moment we are not ready for this. I do not see much benefit for companies to pay for my data. I am a simple person who just love to spend time with my family and work as an ordinary manager. So for now, the digital economy is just an idea.