Quote from: KennyPowers on March 11, 2018, 04:13:38 PM

? How would it work ?
Well there should be some sort of a device connected or implemented in a 3D printer that would serve as imutable proof that work is being done (electricity & material being spent), while also having capability to write transactions or whatever useful things graphic cards do.
An obstacle is: how to prove that useful work is being done (e.g. you're not printing random things and throwing them in the trash afterwards)?

Hi Kenny,

I think a natural case for "useful work" is in distributed computing projects. There are numerous tasks that can be effectively divided into work units and a ready market for entities that want to buy computing power to have those tasks completed. For purposes of this post, I'll talk about a very simple distributed computing use case, OCR text recognition from scanned images of pages. There are businesses that pay to have their files scanned and OCR recognized.

The minimum viable product would be building an OCR engine that works in a pool-mining fashion. However, the work is centralized. So that's not ideal.

Next iteration, allow it so that anyone can submit work projects from any node, they can also set pricing in terms of what they want to pay for each completed work unit. Further iterations could allow them to deploy their own sandboxed OCR engine (and, of course, any number of other distributed computing engines, image recognition, structural analyses, statistical calculations, training ML/DL algos, complex systems modeling, etc.).

I think that would be a useful type of decentralized computing system, there's some work being done in this area and I want to see what can be done to advance this. I've reached out to GPU vendors regarding this for my organization because I believe they'd have an interest in finding "useful work" applications for a mining-type client base. They are making a decent percentage of their income now from cryptocurrency miners, so finding a way to sustain that income through types of mining that is more about useful work as opposed to algos (which can be more easily ASICed) is in their best interest.

I think on Reddit there's a post about using excess mining heat for a greenhouse, I thought that was a pretty cool idea. I don't see "useful work" mining replacing cryptocurrencies, but I do see it as a novel application of distributed computing that can work in a mining-like manner if properly implemented. I am focused on finding models of replacing content monetization that isn't so dependent on advertising. We need new "YouTubes", but we need to be able to pay those that put their lives into earning money off of that platform. Mining a traditional cryptocurrency is only one way that users of a hypothetical decentralized platform can contribute to monetization without actually paying directly. Advertising will probably always be a part of the equation but I believe that multiple revenue models should be explored.

Best regards,
Ben