Enterprise blockchains
Do enterprise blockchains (sadly often mislabeled “distributed ledgers”) matter or are they a fluke of the perpetual misconception of the technology? An unrolled tweetstorm of fairly low technicality. More tweets and storms at @ecoinomia.
Let’s talk about enterprise blockchains for a moment. Or more precisely: inter-enterprise blockchains. Even more precisely: inter-enterprise title transfer logs. Fuck “distributed ledgers”.
Shorter enterprise blockchainers: “Make supply chains transparent again!”
Shorter anti-enterprise blockchainers: “Make supply chains unnecessary again!”
The most pressing issue in supply chain is to account for stuff. Accounting in the sense I mentioned before: who has possession over whose property. Add geography for added complexity: who has possession over whose property where.
The second most pressing issue is, to paraphrase Paul Gauguin: where does stuff come from, where is it now, where is it going? In the real world, stuff is not a virtual apparition created out of thin air, it’s real stuff. Like beer. But not only.
Thanks to global supply networks and outsourcing we can now drink beer brewed in Brazil, shipped to China to be filled into cans made in India from aluminum smelted in Australia using Guinean bauxite. With stuff changing hands a couple dozen times.
Physical stuff comes with neverending complexities. It can get banged up on the way, disappear completely, or be exchanged for something that looks just like it. Any tech stuff comes embedded with a zillion hidden components assembled somewhere on the way.
And even before the first item is shipped, any tech stuff comes with a long intellectual property supply chain embedded in the components it is assembled from. No phone vendor thinks it’s funny if one of those components infringe on someone’s patent. That’s hold-up material.
All of this happens in that vast expanse known as “intergalactic space”. Wait, no: “the market”. Enterprise blockchains are trying to boldly go where no relational database has gone before: to account for all that dark matter stuck between entities.
Of course this has very little to do with the Satoshi model. Physical goods don’t simply fit into an outside-verifiable valid/invalid scheme. And you can’t just wish intermediation away. But regardless…
One of the most urgent needs in supply chain is a global log of orderly transfers between entities. “I’ve taken possession of stuff, and stuff looks ok to me. From now on I’m responsible until you see the next logged handover.”
An important addendum: “Taking possession” not only means taking possession of the physical good, which is necessarily exclusionary, but also of its digital record, which is not.
Just as a side note, to call the world of supply chain and procurement “trustless” would be a euphemism. “Brass knuckles” might be a better term. So this idea that one entity might just collect and control all this info won’t fly.
One other key element of supply chains is the bullwhip effect. The exploding effects downstream of unobserved events upstream. Misalignments between two parties quickly reverberate through the system.
So even if enterprise blockchains don’t implement the Satoshi consensus and disputes cannot be perfectly resolved on-chain, there’s still a point to keeping transfer logs in a system deployed between enterprises and witnessed by other market participants.
All of this stuff is like 30 years old. Economics has barely caught up with it and still mostly considers supply chain a management topic. Industrial engineers otoh only have a very perfunctory understanding of markets. Not to mention computer scientists…
While economists are still discussing the Fisher Body case from 1926 as exhibit #1 for vertical integration, automakers are eagerly pushing in the other direction: the only thing “Porsche” in your 911 is the drive train.
On the other hand, designing an enterprise system (or integral parts thereof) for a non-corporate environment can make an industrial engineer’s head explode. The things you have to pack in when your design goal is Pareto-efficiency and not systems efficiency.
And just to wrap this up, from this vantage point SCSCs (so-called smart contracts) aren’t all that alien. If you move from enterprise-centric to market-centric operations, you’ll want to take your processes with you and write them in a way that they can withstand handovers.
As long as you can keep the different layers of an enterprise (here: operations, finance, legal) apart and see how they interact via transactions (goods flow vs finance flow wrapped in a legal envelope) this isn’t all too hard to dissect. If not, it gets muddled very quickly.
The original tweetstorm appeared on @ecoinomia. The text has been slightly edited and it might still evolve into a real essay. Oliver Beige is an industrial engineer turned economist (PhD Berkeley, MBA Illinois, MSIE Karlsruhe) who is focusing on how technologies like machine learning or blockchain might change the value creation between markets and enterprises.
