Researchhub: GitHub for Science
researchhub.com
researchhub.com
I really don’t see what this adds other than a cryptocurrency component and a leaderboard. I’m not convinced that the missing component in modern science is some form of internet points.
[Disclaimer: I’m a couple decades post-PhD, actively working as a researcher the whole time on all sides : publishing my work, peer reviewing others, and the editorial side helping the whole process work. I am very happy to see progress towards making science work better, but have learned to be super skeptical of startups sniffing opportunities in this space.]
The integration of the cryptocurrency certainly has the potential to veer off in scammy directions, but it also has the potential to introduce incentive design mechanisms to solve some of the issues plaguing academic research at the moment. E.g. incentivizing replication.
But the idea of a Github or Stackoverflow for research that uses technology to crowdsource and mediate the review process (both pre and post publish) and then makes that process (at least the post publish part) open and available for the public to see has a ton of merit and potential value.
I grew up in academia. My parents are both academics, as are my god parents. But for any number of reasons I chose not to go into academia. I used to scoff at the whole idea of an ivory tower when I was on the inside, but having spent a decade and a half on the outside I now understand.
Science is completely walled off from the public. It's happening in institutions that charge more than a house for access, and published in a massive constellation of journals which charge $15 to $30 per paper, or hundreds to thousands of dollars per year for full access.
When you take the cost of a single paper, the fact that papers on any particular topic are often spread across a myriad of journals, and consider the fact that science works in the aggregate then add in the media's dismal scientific reporting track record and is it any wonder we have a crisis of scientific understanding and trust? Once you leave academia, your access to the truths it produces is completely cut off.
We need to find a way to break the journal system. The things they provide to the scientific community - editorial assistance, review, and reputation - really ought to be easily replaced by crowdsourcing and software. We just need to build that software and then convince the scientific community to use it.
The benefits to science and society could be massive, because a public that understands and trusts science means a public more willing to fund science, and a society that has a way to agree upon what is and isn't true.
They also did the hard work on making sure journals will accept papers uploaded to there.
Maybe they don't have a flashy website but I would much much rather use those than upload my work - which I have invested years of my time + thousands or millions of dollars in - to some random site with founders who are just "from San Francisco".
For those trying to "accelerate science" - please try to take even a modicum of time to understand how science is even being done currently and start from there, instead of making up solutions to problems that don't exist.
:%s/science/industry/g
So many of these cryptocurrency-based “solutions” could have this same comment aimed at them. So much time/effort/money spent rolling out web services that don’t even understand the basics of the industry they’re trying to “disrupt”.
I’m not anti-cryptocurrency at all (and have my spread that I support) but this “take x and add a token to motivate people to post x to our site” is getting pretty old.
To give a little background - I'm the CSO of ResearchHub and I come from scientific background. After college I started a PhD program in molecular biology and I've worked in epigenetics and synthetic biology labs. I ended up dropping out after seeing how difficult it was to achieve success as an early-career scientists. A few random stories:
-A friend of mine was told to leave her PhD program with a masters degree because she was unable to replicate a famous result in her field. Three years after she left, the paper she was told to replicate was retracted.
It turns out that >50+% of research is unable to be replicated, so I'm sure this is not unique to her experience.
-A PhD student in biochemistry lab I rotated through was in his 9th year of the program (normal is 5-7), completely out of funding, sleeping on a friends couch, and paying for his experiments out of pocket just to get the publication he needed to graduate with a degree.
Academia is riddled with perverse bibliometric incentives that encourage researchers to partake in behaviors which hurt the ecosystem as a whole. For example, an article shared on HN a few days ago showed how irreproducible papers are cited 153x more than reproducible papers. This is because more interesting results are more likely to receive citations, but also less likely to replicate.
After I left my graduate program I was admitted to medical school. My institution was fantastic because they got med students into the hospital to interview patients during the first week of classes. Without typing a novel, many of our front-line therapies are surprisingly old and ineffective. The waste caused by a broken incentive structure for academic research is felt within clinical medicine.
You are absolutely right that arXiv, bioRxiv, and MedRxiv are fabulous projects that have pushed the culture of open science forward tremendously. Also, should out Open Science Framework - they are amazing.
While these tools help scholars share their work in the open, we hope that ResearchHub and ResearchCoin will help to fill a grab by creating new incentives that can help to loosen the grip of bibliometrics on a scientist's career prospects and funding decisions.
ResearchCoin was included because it allows our community to create a dynamic system for rewarding open scientific publishing/discussion that can change over time. In addition, the ResearchHub DAO to allows scientists to help govern how rewards are distributed. The admittedly quite lofty goal is to create a democratic and evidence-based incentive structure for science.
While it is possible that the cryptocurrency component will make it easy for some to dismiss ResearchHub, we believe crypto is an elegant solution to address some of the root problems in academia.
1. Where does RSC originate? I see that there are 1 billion RSC in existence. Can more be mined? Do you just generate new ones at will? Is this basically a 100% pre-mine of a token?
2. Who owns the RSC that exist? You plan to distribute some proportion of them. Presumably these can be converted to fiat dollars at some point. Supposing they become valuable, can you just sell off the 99% or so that you haven't freely distributed? What keeps the DAO acting in the community's interest?
3. Popularity is a poor indicator to choose to optimize for if you're looking to incentivize high quality. Even if you were able to build a rock-solid system free of bot accounts and manipulation, which is not your current system, the research and summaries receiving the most RSC would likely be clickbaity papers or those which can appeal to the widest audience. Do you have a way to address these issues?
4. What gives RSC a real-world value?
5. How do you ensure the right person is rewarded RSC? If I upload someone else's paper and get a reward, that doesn't seem right. Especially if a few months down the road they wanted to upload the paper themselves to profit in RSC from it. But presumably you would discourage duplicates, so they couldn't re-upload even if they tried. Similarly, how would incentives be structured for collaborative research? 3+ authors to a paper is not uncommon. It seems like the current system is just to reward the uploader, whoever they may be, like earning a "finder's fee". But then you just end up rewarding whoever can scrape and upload papers the fastest rather than those who put in the work to produce quality research.
I welcome the idea of a community built around open and accessible scientific research. But when you add financial incentives, especially when they're this easy to game and with a cryptocurrency entirely controlled by some opaque entity it raises some major red flags. Hope you have the time to answer some of the above.
1. Yes, RSC is 100% pre-mined and earmarked for distribution to the community, team, founders, and ResearchHub corporate entity. We plan to distribute about 5-6% of the total RSC supply per year to the community as rewards for posting science on the platform.
2. In the future, if we run out of RSC and need to add to the total supply the ResearchHub DAO will be able to vote to mint new tokens. There are two mechanisms at play that keep the DAO honest:
* An open source codebase - If the ResearchHub team ever goes rogue and fails to act as stewards of the community, the community will be able to fork our codebase and manage a new version of ResearchHub how they see fit. See the Steem vs. Hive example to see how open-sourcing can give community an option to exit: https://hackernoon.com/inside-trons-steem-takeover-attempt-a...
* RSC holding represent ownership in the DAO. For every coin held, an individual wallet gets one vote. This means that the biggest hodlers of RSC will have the most influence.
On the surface the concept of more coins = more influence might seem like a bad idea, but we believe that carefully distributing RSC into the hands of the scientists and community members who care the most about our mission will result in the best governance.
Regarding guarantees against the team dumping tokens - we currently have vesting smart contracts for employee token grants and would be open to enacting other measures to ensure a healthy market for RSC if requested by the community.
3. I wrote a little bit about this in another response. Let me copy and paste here:
The way coins are awarded to posts can change over time via a democratic process facilitated by a DAO. In theory, this could create the infrastructure for the scientific community to self-govern how their work is assessed and rewarded.
Perhaps a 1-10 peer review rating would provide a better signal than upvotes? Or maybe basic research deserves 1.5x rewards because it helps to fertilize the landscape of translational research. If the community wants to make changes to Researchhub's incentives, we will be able to iterate over time until we find a formula that properly encourages the behaviors that benefit science the most.
4. As of today RSC gives its holdlers the ability to vote in the ResearchHub DAO and participate in the governance of the network. Within ResearchHub you can tip users for their posts by "supporting" their content with RSC.
5. We’re currently building out a feature to help address this. Any scientist who has authored a paper shared within Researchhub will have a pot of RSC waiting for them upon sign up. It is in our best interest to properly credit reward the authors of manuscripts because in an ideal world they would have a larger influence in the ResearchHub DAO compared to the person who simply posted a link to their paper in our forum.
"While it is possible that the cryptocurrency component will make it easy for some to dismiss ResearchHub, we believe crypto is an elegant solution to address some of the root problems in academia."
so
"let me answer or non-answer these questions in multiple separate responses instead on the website I'm running that is definitely not a scam"
I really don't understand how it helps research, can you please elaborate.
Do you mean researchers get paid in a coin, which they can sell, which will fund further research? If the coins became super valuable wouldn't this simply do the opposite: incentivize cheap/easy papers about popular/meme subjects in order to get upvotes?
Perhaps a 1-10 peer review rating would provide a better signal than upvotes? Or maybe basic research deserves 1.5x rewards because it helps to fertilize the landscape of translational research.
If the community wants to make changes to Researchhub's incentives, we will be able to iterate over time until we find a formula that properly encourages the behaviors that benefit science the most.
To speak to your example - if the v1 incentive structure encourages cheap/easy/meme science, we would be able to recognize this and make improvements to hopefully achieve better results.
This is an improvement over the current static system which values bibliometrics because it has the ability to adapt and improve over time.
Doesn't this already happen? Science chose bibliometrics as its God. Making a new God out of crypto and crowdsourcing doesn't feel like it addresses the root concern, which is human bias in the scientific method. It sounds like the system you're divising is just a new way for humans to bungle things with bias (no offense intended). I think you're on to something with trying to develop a new method, but perhaps we should figure out how to remove the bias before involving money and human heuristics?
Now, when you talk about solutions it gets really fuzzy. There're very specific problems that exist and you need to create a well-defined incentive. As a graduate student in bioinformatics, my main struggle is the source code. People either don't share any at all, or share something you can't possibly run.
> we believe crypto is an elegant solution to address some of the root problems in academia.
So what are the root problems? First off, the code isn't written by the software engineers, but rather by the biologists who learned how to write some R or Python scripts. Even the CS graduates who switched to bioinformatics have no idea how to develop quality software.
Next, the universities often have their own hardware for computation(HPC). When running on HPC, you often use the pre-installed modules. For example, my university's HPC doesn't even allow running Docker. So in order to make my code runnable somewhere else, I need to put in an incredible amount of effort. Would I do it for some crypto? No.
The stakes are extremely high in the academic job market, competition is fierce and all incumbents are highly skilled and tremendously resourceful.
Are you prepared to maintain the integrity of ResearchCoin against all sorts of attempts to gamble the system, or even adversarial attacks by those whose careers depend on it?
And this is where you list the root problems in academia and show how yet another virtual token solves them, and this helps people to see its value, and not dismiss it easily.
Oh, wait. You don't. Not here, not on your website. I mean, you do mention that "Further details about ResearchCoin can also be found on the ResearchHub Notion page.", but don't even link to that page.
Additionally, it allows tools like https://plaudit.pub to add functionality like allowing people with a browser extension to endorse the articles. (Disclosure: I volunteer for Plaudit - hit me up if you want to integrate it into ResearchHub natively!)
Although the DOI is most important, there is other metadata you can expose. For more info, see: https://scholar.google.com/intl/en/scholar/inclusion.html#in...
yeah, but they don't use crypto :)
(I'm being sarcastic here if it wasn't clear)
Academia is quite thoroughly broken, everyone in it hates it. The reality is with the existing publishing model you do upload your precious work to some random site. There it is judged by faceless editors, and sold for profit if they deign to publish it, after you have done professional level editing and graphic design for free. And you do this only because of a consensual hallucination that this is not only essential but the true measure of one’s worth as a scientist.
* Does it use git?
* Does it have any version control?
* Does it have community contributions to projects?
* Does it let people open issues on projects?
After clicking around for a bit, I'm not sure what parallels can be drawn between this site and github. Almost none of the core features of GitHub exist here. The site seems to be a PDF file upload site with a comment section.
Also, GitHub doesn’t have a coin system.
I'm not sure I understand this.
And double extra bonus points if you can hide other people’s work from view if you don’t agree with it!
Science can't work if unpopular and controversial results are not given fair consideration. By all means, bad methodologies and conclusions that do not follow from the premises should be picked to pieces, but we can do without downvote campaigns and snarky "Big yikes, sweaty, don't you know that has been debunked already?" comments.
Edit: thanks for the ResearchCoin, kind stranger!
My name's Pat and I'm on the ResearchHub team. Thanks for taking a look at our site and sharing these questions.
"Github for science" is a phrase that describes our long term vision for what ResearchHub can become.
As a poster below mentioned, today's product could be much better described as Reddit for science. It is a v1 designed to test a few hypotheses and bootstrap a community of early adopters who believe in our mission of open scientific communication.
Overtime we plan to add more Github-esque (for lack of a better word) features. Next in the pipeline is a science-specific collaborative text editor that will help teams of scientists publish any kind of research output directly to ResearchHub.
We are still in very early days and have not yet found product market fit. The more critical thinkers we have the better, so if you have any interest in helping us improve ResearchHub I can share a link to our community slack channel.
Thanks again for your interest in our project!
Before my company was acquired (https://assembl.net) I was building something quite similar. I went to many such conferences to meet with my “target demographic,” and I found many of my preconceptions about the future of open science were quite incorrect. I think I attended more than 10 other conferences, and I always found that the problems that needed to be solved were more social and legal than technological (though technology played a crucial role in glueing these pieces together).
I also know many people who tried to do something like what you’re doing. For example, I went through the CV Labs incubator with the guys at https://orvium.io.
Now I’m working on https://intpub.org and https://liras.org.
The deeper problems, I’ve found, are much less “sexy” because they’re not as technological. The fixes need to come in the form of improved licensing models, better publication of pre-publication results, etc. I’ve spent a lot of time thinking about these problems, so I’d be interested to talk and hear where you’re at. If you’d like some sober counsel I’d be glad to provide it.
Personally, I think it means that the site is an attempt to centralize something that is currently decentralized.
with that definition, almost half the internet would be "GitHub for...". at some point, you have to realize that it's just marketing BS and they have almost nothing in common with it.
This is a project which at face value wants a piece of the pie in the publishing market, not something to improve the quality and funding of the scientific community.
As a scientist, I would like to know this project doesn't exist.
Can you imagine how well this will turn out?
This is a terrible idea and you should feel bad.
> This is a terrible idea and you should feel bad.
Especially considering this is exactly how science works now. You publish, get a ton of citations, and get bigger grants.
Source: 20 years experience in competitive research areas.
The research is built on personal relationships and trust. If you work on a niche problem, you already know everyone else who works in the field. You don't really need some crypto-coin to build a reputation, that's what citations are for.
And I don't think it requires a service, and certainly not a Silicon Valley startup. Labs can just put their papers up on their web sites, and make hyperlinks to those they cite, review or otherwise use in their work. Wasn't that the whole point of the web in the first place?
Semantic web can be used to make the metadata machine readable, and PGP signatures can authenticate the peer reviews. These crusty old technologies are perfect for the task because it's the same crusty old researchers that came up with them that are going to use them.
I followed on Twitter multiple debates between paper authors and their peers, most of them very valuable and I feel that they outcome of those conversations are somehow lost in the Twitter stream, not easy reachable and for sure when someone will read that paper will not find those conversations.
One feedback I have is that I don't think science should be voted or the results be presented based on some kind of popularity.
The results of science sometimes gives us an answer we don't like or don't want. But that does not make it (the answer) less valuable or less true.
Chrome with no adblocker shows that it does 111 requests loading 5.6 MB of resources for the about page. 5 of those are Javascript.
Anyone can technically start their own journal as well. The only thing that has to happen is for the journal to gain the trust of the scientific community over time (which, again, is mostly decentralized). Anyone can build a university. All it has to do is to build trust.
No coin or website will make that more decentralized.
> This specification describes SciFed, a standard for federation of scientific activities and content using ActivityPub. It is intended to be used in the context of the Activity Vocabulary and provides a vocabulary for activity types needed in the context of scientific exchanges.
It is from 2019, and probably from the OLKi project: https://framagit.org/synalp/olki/olki
Don't know if it is still active, but the idea to use the Fediverse for Open Science interop is appealing.
There're several incentives to make your research as obscure and unreproducible as possible. First, you want to keep most of the details private to your lab. Why would you help other labs? So that they can make a discovery before you? No way!
Next, suppose it was easy to re-run your whole data analysis pipeline outside your lab. So there's someone you don't even know who tries it, and finds and tiny bug in your code that invalidates your results. Now you have to publish a correction, or even retract. You don't want that happening ever. So you publish some undocumented and unrunnable code so that nobody would even bother reproducing.
Basically, if you provide a software mediated review system (both pre and post publish) and a reputation system, you can effectively replace most of what the journals provide. The hard part is getting the systematic buy in. If you can get that buy in though, the benefits would be huge.
Imagine being able to solve the file drawer problem by incentivizing all publishing, not just novel publishing. Imagine being able to automate literature reviews with NLP and some machine learning. Imagine wanting to know what the scientific consensus was on any given topic and being able to go to a website where you can ask a question, get the literature review in answer, and then click through to individual papers in order to read the papers, their reviews, and the questions of the public answered by researchers. Imagine what that could do for world wide scientific understanding an acceptance - to have it all happening in the open, no longer mediated by a media incapable of good science reporting.
I really hope this gets traction.
The about page is confusing. It sounds like it is trying to be researchgate, but the title says github for science; there is nothing on there about being for code!
It certainly seems like the medical/scientific field could use something like ResearchHub.
I look forward to seeing how this evolves.
There's literally nothing on the entire site that requires any of this. These are mostly static pages with static text.
My name is Patrick and I'm the part of the ResearchHub team. Thank you for taking a look at our site and sharing your thoughts/feedback!
To give a little background - I'm a former PhD student who dropped out and went to medical school after seeing how difficult it was to achieve success as an early-career academic.
We built ResearchHub as a potential solution to what we perceive as the core problem within academia - the bibliometric-based incentive structure.
Job security and research funding flows to the scientists who are the most successful at maximizing their citations. Following Goodhart's law - citations no longer serve their original purpose as a measure for academic quality because they are now a target for behavior. Academia's reliance on bibliometrics unintentionally encourages behaviors that detract from the scientific ecosystem as a whole.
We included a cryptocurrency and corresponding DAO because we believe crypto is an elegant solution which allows a community to democratically manage a malleable incentive structure. To see how this could work in practice, check out Ethereum improvement proposals (https://eips.ethereum.org/). Ethereum is a dynamic incentive structure that aligns the interests of many parties to accomplish a goal. Via EIPs, the Ethereum community is able to improve the protocol overtime in a decentralized fashion.
Today's ResearchHub is a v1 designed to test a few hypotheses and bootstrap a community of early adopters who believe in our mission of open scientific communication.We are still in very early days and have not yet found product market fit - so we are actively seeking feedback from academics and citizen scientists alike to help us refine our feature set.
Once again, we very much appreciate your time and attention. Thank you for digging into our project and sharing your thoughts on what we hope to accomplish with ResearchHub!
If you have any questions about the project, please feel free to comment here or reach out via email to hello@researchHub.com