Problems we don't want to solve - decentralisation and the free lunch

As is appropriate for this time of year, yesterday my sister started enthusiastically describing a new app by the municipal graveyard maintenance company, which allows users to search for a particular deceased person and pinpoint their grave on a map. On top of that, the app shows information about the date of death and interment, it has a time stamped photo of the grave, AND it can be used to order flowers for the said grave, which uses the photo mechanism to confirm to the user that their transaction was indeed completed successfully. I didn't want to bore her with explaining there's a ton of such apps out there. Some even allow attachments such as photos and videos to be added to the grave site, which is kind of neat. I imagine one could walk through such a multimedially enriched graveyard, scan QR codes or somesuch on each grave, and see fond-ish memories of the ex-people resting there. This is a good idea. I would use it. I've even occasionally thought of starting such a business. At least, there's no shortage of potential customers incoming every single day. It's basically infinitely scalable. What stopped me each time is that I have grave concerns (yup) about the longevity of stored data, especially if the company goes under. This is not a front-end problem. We can make this data presentable, searchable, usable in many ways. It's purely a problem of where and how to store this data.

I believe this is one of the poster-child use cases for decentralised computing. Among all the flopping around and general nonsense accompanying web3, mired in the get-rich-quick fintech schemes, there are actual uses for blockchain technologies which make sense. It looks like the best way to conserve data nowadays is to keep copying it and sharing it over the Internet.

But - why?

This is not a problem of technology. We have IPFS. We have torrents. We have various web3 blockchains just waiting to be used by something (anything!) mainstream.

It's also a poster-child case for a problem we, as an entity tentatively called a "civilisation" just don't want to bother solving. It's a minor fascination of mine: there are obviously so many things which surround us which could be made better, but aren't. Usually, it's because people don't agree on thežree aspects of the solution:

• Priorities - should it be solved now?

• Methods - how should we solve it?

• Responsibility - who should solve it?

For example, the issue of decentralised data storage. It's pretty obvious that the concept has huge potential benefits, such as preserving data which shouldn't dissapear when companies go under, or which isn't monetisable as such, but would be useful to keep around.

Relating this to web3, for me entire point of web3 is to run apps that are dissasociated from the messy dangers of the real world. It is a sort of an overlay network on top of Internet, not really reliant on IP addresses, routing tables and DNS. Web3 has the same issues: how to translate the needs for computing power and data storage to real-world machines, in a way which is both safe and provides incentives.

Filecoin seems to be a working combination of the two: a distributed storage system intertwined with web3 incentives which provide "mining" rewards upon proving certain amount of data (i.e. "space") has been stored by the node for a certain time (hence "proof of space-time"). There is no free lunch, at tthe bottom line, a real person needs to pay, no matter how, for the hardware and energy, and they need compensation.

Getting back to the main point: why isn't Filecoin more popular? Or IPFS, for that matter? It's certainly possible and within reach of today's infrastructure to ship one of those on each new laptop sold, where the user just has to enter their credentials maintained by a similarly decentralised identity provider. I can't imagine nowadays NOT connecting my mobile phone and Chrome to Google, because doing that saves me so much effort later. It's certainly possible to have a decentralised ID provider which is just as popular, and services which just naturaly cooperate with it.

Why isn't that particular problem being solved?

• Priorities: there are arguably easier ways for hardware manufacturers to earn money. Enabling end-users to help the community and possibly earn some fees just isn't on anyone's priority list.

• Methods: even if they agree there's a need and it should be solved now, they will certainly rather try to get incompatible walled garden systems in place, instead of completely decentralised solutions, because they hope to get some money out of it.

• Responsibility: today's decentralised systems usually carry a certain dose of anonimity of operation. IPFS for example doesn't even has the concept of authorship / ownership of files. The files just exist on the system, and the system doesn't care where they came from or what they contain. Of course, this could be misused to store various types of illegal material, and no manufacturer ever will want to take on even the vague possibility that they might be the cause of illegal data being stored on their clients' systems.

I think the last point is the most significant at this time.

As the civilisation at large has apparently emphasized the desire for increasing safety, there's probably no way such a system will be widely deployed without a strong way of authenticating data provenance / ownership.

There's only one problem: people who know how to work on decentralised systems don't really want to do it that way, and people who want this type of system don't know how to make it. It won't be solved by IBMs of this world.

Thus, this is a problem that, at this time, our civilisation has marked as "will not fix."