OR: Decentralized Exploitation, a cautionary tail about Blockchain.
Okay, the title is a little (over) dramatic but hear me out- there is reason to be concerned and thoughtful in our actions. A little over a year ago I stopped working on a startup called Bazzta, we were one of the many early-stage supply chain companies touting blockchain as the solution to proving the providence of commodity goods. It started as an attempt to identify problems in the fashion supply chain and when that proved too opaque, we pivoted to coffee, then to honey (I still believe honey was a good idea) before finally closing up shop. We don’t all make it.
The goal was to inform the consumer about the origin of their goods, and we believed that once informed, they would make better choices. Like most blockchain supply chain startups, we believed we could create a more fair environment for the farmers, that consumers would tip through blockchain, that things would get better for everyone.
There are many startups like this, mostly in the United States or Europe, which promise to help farmers and producers in other parts of the world. The sector has become rather hot and some of the largest companies in the world are now involved. One project with significant traction is IBM’s Food Trust which promises “A new era for the world’s food supply.” Fantastic. But there is another quote from IBM’s Food Trust that should plant a seed of doubt in your mind.
Where data becomes the world’s most essential ingredient. Think about that for a moment.
Two years ago when I started working on Bazzta I was in talks with several European based startups about pooling resources and trying to attract investors. Through them, I found a particular company in Africa that was building machine learning-based supply chain technology for coffee. Their setup was ingenious: coffee farmers would input their beans into the machine, the machine would grade them in real-time using machine learning, and then automatically pay out the farmers with cryptocurrency on the spot.
The system was impressive, to say the least, and it seemed, at least in my mind, the promise of Blockchain fulfilled. The system tracked the providence of coffee back to nearly the bean, machine learning built powerful models about the harvest, and farmers were paid out for their crop on the spot and could get a higher price for their coffee thanks to the efficiencies gained by the system. It seemed like a truly fair, equitable, and better system.
Or was it?
As I continued working on Bazzta I always had this machine learning system in mind. We could build similar machines for nearly any kind of crop, work directly with the agricultural cooperatives. We could save farmers from the expensive, exploitative bridge loans they take out to cover the time between harvest and market. I had even more exotic ideas- micro commodity markets where farmers could pre-sell on a global decentralized exchange shares in future harvests, hipsters in San Fransisco could long bespoke coffee strains on decentralized exchanges. Coffee drinkers in Brooklyn could buy and sell fair-trade coffee futures from hipster cafes where tips were sent directly to the farmer. Farmers could take loans out against aficionados in Europe who would back purchases of new equipment in exchange for shares in the crop. MakerDao style stable-coins backed by commodity prices. Everything was possible.
At SXSW I spoke with IBM about Food Trust. they had seen another angle on the problem- food recalls caused by salmonella or E.Coli on leafy vegetables took weeks to track down the source of contamination. Even once the contamination source was found and eliminated, on average it took over ten years for the affected product category to recover, costing tens if not hundreds of millions in losses for producers. At the time apparently Walmart was on board and directing its producers for leafy vegetables to switch over to Food Trust. It was happening.
As I studied the market to better understand the coffee business, I ended up speaking with Serra Trading Co. from the Blue Mountains in Jamaica who were interested in blockchain solutions. In our conversations, they mentioned the African company I had found earlier, and to my surprise, they expressed serious doubts: “Transparency that’s good for everyone” is the companies slogan but Serra Trading Co. wasn’t buying it. After evaluating the solution, the machine learning, the data pipeline, they had come to realize what IBM had realized: it is not the coffee that is the product anymore, it the data about the coffee that is the new unexploited natural resource.
Think about it.
A single coffee bean is almost worthless alone, it is only in bags of thousands of other beans that they have any value. Likewise, the data about a single coffee bean is worth so little that a farmer is incapable of extracting any value from it. Even cooperatives that represent scores of farmers are limited in their ability to extract value from the data of a single bean.
The majority of coffee (70%) is produced by smallholder farmers (5 acres or less). How can what they produce be considered a commodity? So little value is placed on the farmer and the coffee they produce. I guess if you don’t value the farmer & their crop it only makes sense that you believe they shouldn’t own their data as well. But we know this unfair and untrue.
— Andrea Johnson President of JAWiC (Jamaican women in coffee) www.womenincoffee.org
Of course, if you can build a supply chain pipeline that can capture the data about every bean, suddenly everything is different. With enough data you can create a comprehensive picture of the heath of harvests, the effect of fertilizers and farming methods, you can understand rainfall, climate change, and yields. You can look into the past and predict the future, correlate growing conditions, identify and eliminate inefficiencies and standardize quality. Most importantly, however, you can create an entirely new resource: data, and the more of it you collect, the more valuable it becomes.
This would be fine in theory, but in speaking with Andrea Johnson, President of JAWiC (Jamaican women in coffee) I realized the data collected doesn’t belong to the farmers. It belongs to the owners of supply chain software. From an investor’s point of view, it sounds brilliant, but to some coffee farmers, it suggests something different. If data is the new product, the new commodity, then the farmers are still only being paid for the bean. Blockchain-powered supply chain startups like our own were promising farmers marginal increases in value while simultaneously extracting data as entirely new natural resource.
When I realized this myself, suddenly the application of blockchain technology for good looked just like the exploitation of developing economies of the past. A pittance paid for rare-earth metals, for lithium, oil, for cocoa beans, cotton, human labor or palm oil, things found in abundance elsewhere, but “valuable” only once extracted and processed by us, for us.
Coffee prices are based on an outdated model. For some reason, the cost of labor is the elephant in the room. No one “knows” the true cost of production. Why we are looking at solutions that allow consumers to tip farmers instead of paying a fair price for their product?
— Andrea Johnson President of JAWiC (Jamaican women in coffee) www.womenincoffee.org
Of course, true capitalists won’t mind this imbalance, the “value add” seems entirely justified. Rare earth minerals mined in Malawi have very little use or value until processed into smartphones in China, (and then resold back in Malawi). The data from a single coffee bean also holds little value to a farmer in Ethiopia, so what’s wrong with repackaging it with a billion other beans and selling the data on to others?
Data however, unlike rare-earth metals, is forever. Data only grows in value the larger the data set becomes. While theoretically Malawi might one day process and refine their rare earth metals into smartphones themselves, the data extracted at nearly no cost from coffee farmers will never be returned to them except for at a significant cost. Every year that the private data set grows and is aggregated with more locations is another year of first-mover, monopolistic advantage that local markets will forever struggle to compete with.
As companies like IBM rack up more success in supply chain tracking, the issue of data ownership will only grow. The monopoly of data as a product is an important value proposition of businesses working in the blockchain supply chain space. The education of producers to the value of their data is critical to prevent a new kind of “digital colonialism” where we pretend to fairly compensate producers. We are being dishonest and exploitative when we pay producers for only product they currently understand but extract from them the true wealth and value of their harvest for a pittance.
One solution that I find encouraging is the growth of locally developed supply chain solutions and groups such as Women In Coffee. One project that I still speak with often is Affogato.co, based in Honduras they are working directly with coffee farmers and cooperatives to develop solutions tailored to their coffee value chain. Decentralized solutions built by local teams are hopefully more likely to build solutions that fairly support and empower producers. Solutions controlled by DAO’s (Decentralized Autonomous Organizations) such as those powered by Aragon could use blockchain to coordinate and extract the value from large datasets and redistribute fairly the profits.
After considering our options at Bazzta, we pivoted to honey before eventually closing up the shop. It was a difficult choice as I was passionate about the subject, but it was clear that building the future without being exploitative would be difficult. Companies in the supply chain space will need to carefully consider whether or not their value proposition for investors comes at the expense, and exploitation, of the producers they claim to help and empower. The real danger at the cusp of freedom is that the same technology we bring to liberate, might prove dangerously tempting exploit as well.