Twitter | Search | |
(((Lukas Kawerau)))
Super fascinating post on "scaling knowledge" at AirBnB via I wonder...is this not how science should work? Like as a process? One repo, every article a pull request discussed by peers before it is merged into the body of knowledge?
Reply Retweet Like More
(((Lukas Kawerau))) Dec 19
Replying to @LukasKawerau
Maybe we should just do away with this whole publishing in journals business. Set up repos for every (sub) field and have people submit pull requests to add their knowledge to the master branch of that field.
Reply Retweet Like
(((Lukas Kawerau))) Dec 19
Replying to @LukasKawerau
Reputation (which is oh so important in science) is inherent: number of accepted pull requests. Add in continuous integration services to make automated tests of the data pipeline and analysis process.
Reply Retweet Like
(((Lukas Kawerau))) Dec 19
Replying to @LukasKawerau
Collaborative process where discussion of the PR is used to improve and enhance the end result. You could even slap on pre-registration by requiring the first commit to the PR to contain what's to be investigated.
Reply Retweet Like
(((Lukas Kawerau))) Dec 19
Replying to @LukasKawerau
Only thing I'm still struggling with is anonymity. Maybe anonymous profiles and the PR needs to contain an encrypted file to prove ownership after PR is accepted? Or do away with anonymity? That's the main stumbling block I think.
Reply Retweet Like
Tiago Forte Dec 19
Replying to @LukasKawerau
Why do you need anonymity?
Reply Retweet Like
(((Lukas Kawerau))) Dec 19
Replying to @fortelabs
For the reason what we have double-blind (and hopefully soon tripple-blind) peer review: so that the reputation of the person submitting has no influence on acceptance. That's the difference to programming: code doesn't run, don't include it, no reputation in merging.
Reply Retweet Like
(((Lukas Kawerau))) Dec 19
Replying to @fortelabs
But science is not clear-cut, so we need to try to take that out of the equation as much as possible.
Reply Retweet Like
(((Lukas Kawerau))) Dec 19
Replying to @fortelabs
But maybe we don't need it? I'm not entirely sure, but the experience of the sciences shows that that is a useful feature of the current process, I think.
Reply Retweet Like
(((Lukas Kawerau))) Dec 19
Replying to @LukasKawerau
To be quite honest, I'm very excited by this mental model. You could have issues and discussions of these issues that are referenced in commits, PRs, etc. which would be very helpful. I think there's a lot here that could be very useful.
Reply Retweet Like
Tiago Forte Dec 19
Replying to @LukasKawerau
Interesting. I consider reputation to be such a useful filter but I could see it obscuring real science
Reply Retweet Like
(((Lukas Kawerau))) Dec 19
Replying to @fortelabs
Oh in general it totally is. But here's one example for why it's bad in science: "Our analysis shows that single-blind reviewing confers a significant advantage to papers with famous authors and authors from high-prestige institutions."
Reply Retweet Like
Tiago Forte Dec 19
Replying to @LukasKawerau
Makes sense. In my world of bloggers reputation is everything, because we're creating primarily entertainment, not truth
Reply Retweet Like
(((Lukas Kawerau))) Dec 19
Replying to @fortelabs
Oh for scientists reputation is also everything, probably even more than for bloggers (🙄). But since we are, as you point out, in the business of creating truth, we have to figure out ways to take that out of the process as much as possible.
Reply Retweet Like