Many blockchain projects like Filecoin, Hypernet and Golem are working on decentralized forms of compute and storage. Don’t dismiss them — they might matter more than you think.
The obvious flaws. At first glance, these projects seem like intellectual cryptonanism. “Let’s take advantage of unused resources” is a compelling narrative. But it pales in comparison to the efficiency gains a centralized system (NVIDIA, AWS, etc) can achieve with relentless optimization.
Decentralization is a serious tax on a product quality. And it isn’t an interesting feature to most app developers (unless you’re doing something illegal). People want a simple, cheap and fast API. It seems unlikely the Filecoin will beat S3 at cost, or Hypernet will beat Google at TPU’s. Or is it?
Counterpoint #1: Miners as companies. Bitmain takes advantage of centralization to mine Bitcoin efficiently. Imagine someone doing the same, for Filecoin. Or Golem. You’d raise money, build a small datacenter, deploy Golem in it, and directly start competing with AWS. One day, you might make your own chips. (One twist is that due to the open nature of the protocol, you’ll need to develop alternate moats. Switching cost shouldn’t be a factor if compute jobs can be easily re-routed to a competitor.)
Think less “SETI at home”, more “AWS startup”. This may be the only way to bootstrap a serious competitor to the incumbent cloud services.
Counterpoint #2: Speculation. In almost every commodity, the speculation market is much larger than the underlying asset. It may be that Filecoin is only used for small contraband use-cases (e.g. sci-hub), but speculating about it becomes a major activity. (This hasn’t emerged to date with AWS due to the friction involved in setting up derivative markets.)
Counterpoint #3: The Blocknet. Instapaper shut down access to EU users this week since it isn’t GDPR compliant. If GDPR is the beginning of “the local web” (where each country has it’s own Chinese firewall) then decentralized compute becomes much more interesting. To date, we’ve enjoyed the Internet as a global network. If that changes to a country-specific service, there will be consumer demand for “foreign goods”. Some of that will be met with kludgy VPN. But if built properly, these cryptocurrencies stand to play a dramatic role in making the Internet global (again). On the Blocknet, you’ll be able to access any website you want. Regardless of compliance with your local country law. No VPNs. No popups. It Just Works. Technology saves society, just in the nick of time! Which leads me to my final point.
A curious coincidence. The world is going through an anti-globalization trend. Refugees rejected from countries. US/China trade wars. People seem to be developing “us versus them” narratives. Watching Zuckerberg’s EU parliament testimony, I was struck by how much anger there was towards Mark. Part of felt directed at him as an “outsider”.
It’s really weird that at the same time a technology emerged (from nowhere, authored by an unknown creator) that provides a counter-balance. Cryptocurrency is less about decentralization, more about reglobalization. If nobody is in charge, everybody is in charge.
Is the timing a coincidence? Or are these factors intertwined? To what extent is the religious zeal around cryptocurrency a political movement only possible in the current era? Could it have succeeded if it arrived 20 years earlier? How would the world look if it was never discovered? Let me know your thoughts on Twitter.
 Like Bitcoin and BCH, I suspect the largest “miners” will just try to fork the network in their favor. The biggest lie about cryptocurrencies today is that they don’t really decentralize. All incentives will lead to capitalistic centralization. One day we’ll have some type of proof that is truly dispersed by nature. My two favorite candidates are (a) a coin based on DNA or (b) true proof of location.
Thank you to Matt Huang and Avichal Garg for reviewing this post.