Ghost Work: How to Stop Silicon Valley from Building a New Global Underclass by Mary L. Gray and Siddharth Suri

Ghost Work: How to Stop Silicon Valley from Building a New Global Underclass by Mary L. Gray and Siddharth Suri tackles a new class of worker: those who exist to bridge the gap between what Artificial Intelligence systems can and can not do. As Gray and Suri note, "the great paradox of automation is that the desire to eliminate human labor always generates new tasks for humans." In other words, there's a gray area between the robots taking over and human labor, and this is it.

These workers are the ones that decide if a picture that was flagged is obscene or not (is that a thumb, or something else?); if something constitutes hate speech when the algorithm can't tell; if a person posing as an Uber driver is really a registered Uber driver, etc. Ghost work is the bridge between AI an an automated future; and as Gray and Suri attest, these jobs complicate the the dominant story of humans being replaced by robots. Yet, these workers are generally invisible, poorly paid and have few protections. (Of course, as they point out, "according to the U.S. Department of Labor's Bureau of Labor Statistics, only 52 percent of today's employers sponsor workplace benefits of any kind.")

The book is the culmination of a five year anthropological study of workers in the US and India using four "ghost work" platforms to make a living: Amazon's Mechanical Turk (MTurk); LeadGenius; Microsoft's lUniversal Human Relevance System (UHRS); and Amara.org.

Ghost Work is, as Gray and Suri show, just the latest in of of 50 years of trends in employment in the US with employers seeking to avoid the employment laws and classifications that would require them to contribute to social security, employment taxes, or other benefits like health insurance. (Uber has said this week that classifying their workers as employees rather than contractors would increase their labor costs by 20 to 30 percent). Microsoft invented the permatemp, contractors who were classified as permanently temporary workers so as to avoid paying for these things yet kept on for years, they note. Tech has been exploiting this labor classification loophole ever since. It is widely reported that more than 50 percent of Google's workforce is contract labor, and many of Google contractors live in their cars in the parking lot of headquarters, not able to find or afford housing in the surrounding community.

Ghost work, like capitalism at large, has a number of myths that make the magic possible, particularly around the flexibility and autonomy these jobs are supposed to provide. The flexibility in scheduling these jobs really involves being on call 24 hours a day – hyper-vigilance is required to get new jobs before others snatch them up. All the risk and investment of this type of work lies with the employee – when there are problems with the technology behind the platform, generally workers don't get paid.

There are numerous stories here of people working within the API driven world of Ghost Work across the United States and India, and an exploration of the dimensions of this work, good and bad. For one woman in a rural village in India, whose take home pay of $350 a month makes her the richest person in the village, ghost work is a fantastic alternative to other local options, but that doesn't mean it is a life path or even a long-term option. AI creates and disappears jobs as quickly as algorithms are deployed and improved.

Given these issues, Gray and Suri have a two groups of suggestions on how to fix ghost work and make it a sustainable class of employment, from both a technical and social perspective.

On the technical side, ghost workers need more tools for collaboration, and a "digital watercooler" where they can discuss work issues with their fellow contractors. Shared workspaces issues by companies could also provide physical collaborative spaces for workers and decrease feelings of disconnectedness and loneliness. They also suggest that employers could create "flash teams" that work together on vexing piecework projects. Finally, they'd like to see "portable reputation systems" whereby ghost workers could share their portfolios of work and demonstrate their range of abilities across different proprietary platforms where they work, to make it easier for employers to find the talent they need.

There are also social fixes to make the work of ghost workers more tolerable. Importantly, they suggest, employers need to build empathy for workers, and institute a "good work code" for supply chains that involve ghost workers. This could involve pledging minimum weekly payments, non-discrimination policies, or agreeing to not charge workers fees to receive the money they’ve earned. We also need better employment classifications for ghost workers and gig workers everywhere; and a safety net for future workers as this realm of work continues to grow exponentially – the authors argue that the gig economy makes the strongest case yet for universal healthcare, public education, paid family leave, and municipal co-working space. Finally, unions and platform cooperatives could work as a countervailing force to hold platforms accountable – if they existed.

Consumer action is the final suggestion as a "fix for us all." Just as consumers have won victories for workers after the terrible fires in Bangladesh garment factories through the Bangladesh Accord, and likewise for tomato growers through the Fair Food Program many fast food chains have adapted, consumers could also demand reasonably decent conditions for workers from the tech companies that use ghost work. Whether or not they will remains to be seen.

This is a wordy book and could have used some tighter editing; some of the concepts are repeated ad naseum. Yet it is an excellent look at the horizon of gig work and where automation meets humanity, and where it could go from here. These contingent labor markets have seen sharp growth, and are only continuing to grow as a share of overall work.