Technology behind Bitcoin could aid science, report says | Hacker News Reader

Technology behind Bitcoin could aid science, report says

physicstoday.scitation.org

13 pointsdigital5514 comments

I cannot imagine someone looking at the problems listed and coming up with a blockchain as the ideal solution. I'm with Mr. Himmelstein from the article, this proposal should be avoided.

The last line is actually explicit that the whole thing is about desperately trying to find a nail for the blockchain hammer:

> Digital Science is offering a $30 000 catalyst grant for an existing blockchain technology or a new idea that can be applied to academic publishing.

Well, blockchain has exactly the role that peer-to-peer has in Joel (on software)'s architecture astronautics essay. A technology that is fashionable due to one successful application. Charlatans and honest fools try to use it for applications where it makes no sense, the press reports straight-facedly because of dem clicks. It is painfully obvious.

https://www.joelonsoftware.com/2001/04/21/dont-let-architect...

Disclaimer: work for, but don't speak for, Digital Science. Am not involved in the grants.

DS gives out catalyst grants typically twice a year. There's a lot of talk about problems that could be solved with this kind of tech and a grant like this is a bit of a boost to actually try them rather than just talking about them.

Nothing wrong with a little technology push, we'll find out in due time if it's actually a good solution. What's 30k usd after all?

A technology push is better than nothing, but constraining a system to blockchains for no technical reason is like constraining it to be written in perl. You're just making the job harder then it needs to be.

Disclaimer: work for, but don't speak for, Digital Science. Am not involved in the grants.

I'll try and add that as a boilerplate disclaimer on each comment, but to be very clear I am not speaking for DS at all. While I like DS (otherwise I'd work elsewhere), I'm very sceptical of blockchain things.

I think this would be more valid as a criticism if it was the only grant available. DS run, typically, two per year which have no requirements like this. $30k is not nothing but it's also not a huge funding round (nor is it intended to be), it's a kick to get some ideas off the ground

To be honest, I'd be surprised if you went to them with "this is not a blockchain but it solves the exact same problems of distributed trustless consensus with x y z, but more efficiently" and got a bad response.

Sure, there's a lot of hype, but there's also not really that much actually built using the tech. Sounds like a fairly reasonable thing to give a bit of money to people to actually build things.

I think the problem is in trying to force application of "distributed trustless consensus" to way more things than actually require them, not in that there are better distributed-trustless-consensus solutions.

I agree, but I don't know if that applies here. Again, there are grants twice a year without any restrictions like this. If you've got a good solution to a problem in the relevant domain then you're not being excluded by another grant being made available.

Bloom is working on non-academic credit building and scoring.

Hyperledger brings together many great projects and tools which have numerous applications in science and industry.

Is a blockchain necessary? Could we instead just sign JSONLD records with ld-signatures and store them in an eventually or strongly consistent database we all contribute resources to synchronizing and securing?

That's just the centralization or decentralization question.

We can do it all centralized already but we would also all need to trust whoever is hosting this data and trust every single person who has the ability to enter the data.

Less nodes you need to trust the better, in a centralized solution where everyone can contribute you need to be able to trust everyone.

In a decentralized system where everyone can contribute, you don't need to trust anyone but give up benefits of centralization such as speed, performance and usability.

> We can do it all centralized already but we would also all need to trust whoever is hosting this data and trust every single person who has the ability to enter the data.

There are plenty of ways to minimize trust required with traditional cryptography though, this is not all or nothing, we have been doing this since PGP. You can get the overwhelming majority of the benefits with none of the drawbacks.

But how else are you going to hype an ICO with claims about the size of a market and get people who don't understand how blockchains work to give your their BTC/ETH?

> Many academic publishers have had trouble finding a sustainable business model

I would say they have trouble finding an ethical business model. Most big academic publishers are printing money. Reviewers, many times also editors are unpaid while the publishers rack in publication fees from authors and/or subscriptions from readers. I fail to see how blockchain solves this issue.

Disclaimer: work for Digital Science, don't speak for them, all my own thoughts, have started on my beer advent calendar, etc

I think many publishers really are having trouble finding a sustainable long term business model. That doesn't mean many aren't making money now, but there is a complex and difficult transition happening. There's a huge push towards open access, and "preprints" (a concept I can never really understand in a time when they don't actually print things). What can be charged for, where are the margins and most complicated of all: how do they get there from here?

Where should the payment come from to build and host the infrastructure, manage the data quality, do the admin, etc? What out of those should even be done any more?

> I fail to see how blockchain solves this issue.

There are at least a few things I can think of:

* Registering an experiment, like a clinical trial

* Registering results

* Having some evidence you've registered a method and proposal before gathering the results, all data in the open

* Register, in the open, facts or results tied to you for credit when they're used

* More, since I'm not really working on this side of things. I do work with bibliographic data though and I'd love and append only log of publication data people pushed to.

But then if everything was clear, it wouldn't be useful to offer a bit of money to gather ideas and try something out, would it? Worst case, the company loses some money but makes some contacts for people with interesting ideas in the field.

Reply on news.ycombinator.com