I like the idea of a decentralized Internet Archive in theory, but I'm very skeptical about the idea of decentralized (as opposed to distributed) filesystems, a

I like the idea of a decentralized Internet Archive in theory, but I'm very skeptical about the idea of decentralized (as opposed to distributed) filesystems, and tying the whole system to a cryptocurrency seems like this creates additional volatility exactly in a situation where you want stability. Buying hard drives and sticking them in data centers is a very cost-effective, efficient way to store large amounts of data, and with careful design (Internet Archive definitely qualifies) you get amazing durability. The failure rate of hard disks is relatively well-known or "mostly" bounded and predictable, so you can fine-tune your data encoding to get a good tradeoff between storage efficiency, CPU usage, network usage, and durability. With a decentralized system, you're adding intermediaries with some amount of additional risk that is difficult to model. The idea is that these intermediaries have "leftover" storage which is basically free, and Filecoin lets them sell it. However, the risk that these intermediaries lose the data is difficult to model and so you would likely choose a wider encoding (https://en.wikipedia.org/wiki/Erasure_code https://en.wikipedia.org/wiki/Erasure_code) with poorer storage efficiency and higher CPU usage. Meanwhile, the cost of buying storage outright is already quite low, and the cloud providers are already busy selling storage (or using it themselves) in high-IO applications like active databases, and this subsidizes raw storage (with low-I/O requirements). Essentially, if someone in the cloud needs a 2TB drive, the cloud provider can buy an 8TB drive instead, split it in half, and sell it as "2TB + priority I/O" and "6TB + best-effort I/O". Finally, in order to be stable long-term, Filecoin needs to provide enough income to miners so they can be profitable. I don't see a good story for how this can happen. The most profitable thing you can do per-byte is to sell "leftover" storage on existing systems that are not full, but this doesn't give you enough to cover overhead, so you would want to run a large operation. Once you are running a large operation, your costs would be similar to that of a cloud provider, but your customers would be choosing wider encodings with lower storage efficiency. If Filecoin were to turn into a common market for larger storage operations, it would then make sense to vet those operations and develop business relationships with them, because by reducing risk you could increase storage efficiency and reduce costs. There is a de facto way to do this already: several cloud providers implement the S3 API, so you can shop around fairly easily, as long as you are willing to transport the data. Short story is that distributed storage is in a race to the bottom with or without Filecoin, it's already "hyper-competitive", and I'm unconvinced that Filecoin provides benefits that outweight the drawbacks.