Inspired by the 180 websites I will understand 52 academic papers in 52 weeks | Hacker News Reader

Inspired by the 180 websites I will understand 52 academic papers in 52 weeks

swizec.com

115 pointsSwizec52 comments

And in turn, you have inspired me to do something similar. I read (to varying degrees of attentiveness) many, many papers per week, but I never write about them other than to take notes, write questions to the authors etc.

For the next 52 weeks, I'm gonna try to write a layman's explanation (with some amount of appropriate background) about a paper every week. I suspect these will center around my work, but I also suspect the act of thinking about the kind of stuff I do in a more general sense will be a valuable learning experience, and who knows, maybe someone else will learn something too!

Alex! It's awesome that you've decided to do this and you have my support! I'm replying to you because I want to pitch you something: I'm running a free open platform for simplified summaries of scientific papers: http://www.sciencegist.com

If you decide to contribute these layman's explanations to Science Gist, I can guarantee you will earn yourself at least one user of the day award :P

Award: https://twitter.com/ScienceGist/status/367559501349543936/ph...

On a scientist's desk: https://twitter.com/RobBeagrie/status/367918271095242752/pho...

Good luck!

I think you need a little quality control on the tags. What's with all the "keyboard mash" tags, ie. ArxicwxQlRqh? http://sciencegist.com/tags

Fixed. Sorry! I had a recent bout with spammers :/ You can add a gist anonymously if you don't like to expose yourself, but having that functionality is quite dangerous online, for obvious reasons. I don't like captchas, but I'll have to add them soon (for anonymous contributions) if this continues.

I was just reading about spambots filtering without captchas, maybe you'll find it interesting too: http://nedbatchelder.com/text/stopbots.html

A researcher in my field did this (not 52 weeks unfortunately). He basically made his blog into a giant review article [1]. He was a celebrity at the biophysical society conference last year because of all the students that discovered his blog on Google while looking for papers.

[1] http://lettsscience.com/

As somebody who is about to start grad school, blogging about papers publicly sounds like an excellent idea.

At the beginning of my Master's degree, I was quite careful about making notes about every paper I read, but as I gained more background and started focussing on doing on my own work, I'd read papers only enough to get what I needed out of them (fortunately I have a good enough memory that when I needed something I'd already read about, I could often find the paper -- and I was religious about at least saving all the papers I read). Doing it publicly might keep the motivation up.

> As somebody who is about to start grad school, blogging about papers publicly sounds like an excellent idea.

Eh, maybe. Publicly criticizing the work of others in an informal venue where you don't have the benefit of editing/peer review could potentially make you look foolish while also alienating potential future collaborators/colleagues/reviewers.

This is an excellent idea! I would definitely love to read your primers on various subjects. Will you share the URL where you will publish your stuff?

Edit: Ah, okay, I just found your homepage. I look forward to reading your writings!

Any interesting papers on your list you do not mind sharing?

First up is "High-resolution reversible folding of hyperstable RNA tetraloops using molecular dynamics simulations" Chen & Garcia (September 16, 2013, doi: 10.1073/pnas.1309392110 PNAS September 16, 2013) which is a bit of a cheat because I'm teaching it next week, and I know Alan.

The following week will be; "Oral Treatment Targeting the Unfolded Protein Response Prevents Neurodegeneration and Clinical Disease in Prion-Infected Mice" Moreno et al (Sci Transl Med 9 October 2013: Vol. 5, Issue 206, p. 206ra138

And after that something from Don Cleverland (UCSF)'s lab, as I saw him speak not that long ago and he blew my mind.

That being said, I'd also (probably) take requests!

Why don't you make it a business model and monetize laymanizing/summarizing papers by request? Just thinking, you could, you don't need to. If you consider that, you could invite colleagues to do the same on a distributed platform/blog. So that every new scientist can just connect and add value, without getting vendor-locked. Who knows, maybe there is a market for selling the results of "peer-review" ;)

You know what? I wish that you get rich with that! Everbody would benefit from more openness in Science. I hate that many many publishers squeeze scientists out like parasitic worms and also cash on us and other scientist who just want to read the paper. They rip us off in the range of $5-$300 per paper and do that with mostly keeping 100% of the profit.

Their industry damaging business model: A←parasite-publisher→B

Breaking Bad: With an open business model: A←→B←→C←→D …

There could be potential in that idea. Ff it were a few more people and every one covering one specific area with one new paper each week presented, it could make an really interesting grad student brain incubator. I would definitely join in if there were a bit coordination.

It is said that the best way to learn is to teach.

I read a Joel Spolksy article about invalidating bad software patents. He mentioned it was very easy. How about invalidating 30 patents in 30 days ... From each of us?

http://joelonsoftware.com/items/2013/07/22.html

The idea is a bit preposterous. There are tons of papers the author has not the slightest chance of understanding in under a week (considering he has to write about it, too). Going from paleobotany to financial economics to law to number theory in one month?

I suggest you divide your year into quarters or thirds and select a theme for each. Spending three to four months studying papers from a certain subject would give you a better chance at understanding, I think.

Good luck, have fun.

Reading this article, Jennifer's, and the comments here have really inspired me. I think I am going to join you guys. Starting today, I am going to code for 8 hours every day for my full-time startup job. Hopefully, at the end, I will have a much better understanding of how much total time of how much of my life I am wasting working on other people's dreams when instead I could be reading research papers, building websites, invalidating patents, learning and practicing design...

Hey I started doing the exact same thing. Only for CS. Minus the understanding part. I think if I can grasp even 20% of it, then the time spent pays off.

If you want to warm up, start with ACM journal articles. They're still pretty dense but aren't quite as formal.

The problem with many CS papers is they're focused at other experts in the same niche rather than to someone with a general knowledge of CS. ACM (and similar journals, I suspect) articles give similar information but are aimed at a more general audience.

I like the interdisciplinary focus and the commitment to reading science at the source. It seems many people have lost touch with science because it is "too hard to understand," they will "feel stupid," or because science has become a closed community behind pay walls. It would really benefit everyone to have more scientific literacy.

I am striving for something similar with a club, the Academic Journal Club, at my university. I hope to move much of the discussion online so others can benefit from our learning. This is our forum http://ajcutk.com/forum/

This is a wonderful idea: I particularly like the aspect of reading outside of your primary field. It seems that many breakthroughs come from the rare person who can take a concept from one field and apply it to a seemingly unrelated field of which they also have knowledge through some quirk of fate. But of course, it need not be pure chance ("the book on crop tilling was the only book I had while I was stranded on a desert island"), it can instead be self-enforced. Good luck!

Are you planning to reproduce the experiments in every paper too?

Even if he would work full-time on that, it would be impossible to do in most cases. A week is usually not enough time to organize an experiment, especially when you are not experienced in the field.

I do not. While I would really like to be able to do that, the reality is that I just can't. I mean, I have a biologist friend who regularly clones bacteria and implements binary functions in their DNA. Not only do I not know how to hold a pipete, I just don't have the equipment to perform experiments like that.

The goal is to get a functional understanding of as many different topics as possible. Enough to explain things in layman's terms, but I could not possibly become an expert in everything.

At least for some of the Machine Learning and related papers (specially old ones), to reproduce the experiment should be relatively easy.

Reproducing those would also be a big step towards understanding them better.

And also a big step to catch these papers in which the results are not exactly as good as the authors said..

You have inspired me to do something similar. Being a college student, I am leaning new things everyday. Well we all learn new things our whole lives. I am working on a startup website and I have decided to code one feature every 2 days. This way I will be learn a lot on web development. Plus this will help me to make my customers HAPPY ! So its a win win.

Which papers?

And what's the deliverable? That is, how do you measure whether or not you "understand" a paper? The 180 websites speak for themselves. How will we know you have succeeded?

I suggest that you be queried by an expert in each paper's field and given a numerical score 0-100!8-))

How will you select the papers? I would suggest you to read the real groundbreaking papers, but it would still be a difficult task to select these from all fields of science.

I'm going to push commit every single day. And write an article per week.

Is there a easy way to stick the promise and not procrastinate?

Beeminder (https://www.beeminder.com/) and similar, based on "social contract". Basically, you stick to your plan or you pay money.

You can find lots of online and mobile apps that are based on the Jerry Seinfeld Chain. A particular example would be iDoneThis.

This sounds like a really great idea! What are the go to resources for academic papers?

you must be way smarter than me. it can take me months / years to understand papers that are not in my field.

I think there are different levels of understanding.

I can't speak for the original poster, but I'm an extremely pragmatic sort of reader and will move on from material once I "feel" I have a grasp of the concepts involved and can "see where the author(s) are coming from". Others might not be comfortable with saying they understand a paper until they reach a similar level of insight as the authors and perhaps even the level to reproduce the results.

When you receive help acheiving your goal from professionals will you state so on your blog or will you, like Dewalt, avoid mentioning that not-quite-flattering fact?

Thanks to the HN detectives some weird jumps in proficiency were found in her code...

Why wouldn't I? I think it's more flattering to ask for help than to splat your head against a wall.

very cool!

meh

meh.

don't forget to get enough exercise and pauses.

You did not get it. No kapish.

This is all lame, building websites, reading papers. These activities are good in their own stead but they don't help the world or achive something spectacular. Why do people want such quick returns. I give you 2 decades, win me a Noble prize.

> they don't help the world or achive something spectacular

Sure they do.

What if I become really good at summarizing papers for laymen? What if I then start a publishing house for good laymen explanations of science? What if this then leads to the abolishment of all those stupid "Caffeine will give you cancer lololol" articles your mum reads?

Or maybe I'll just learn a lot and use it as an advantage somewhere else.

Or it might inspire people to read more base sources to form informed opinions about things you care about.

And I'm almost certainly not smart and persistent enough for a Nobel Prize.

If you do become really good at summarizing papers, you are welcome to join our (shameless plug of a free and open source project) Science Gist (http://www.sciencegist.com) community. It's early days for this project, but I believe it has a huge potential. Also, since you're living in the same city as I am, we should get coffee/tea sometime soon and talk science.

Or join forces with https://medivizor.com/

Actually what i want to point out is, what's the rush?

Reading papers will be your first step in winning a Nobel Prize (unless, of course, you’re aiming for some other, specifically noble, prize :)).

How do you expect to do anything spectacular if you don't put in effort to get good at something?

I am not trying to be snarky, but do you do anything that helps the world and/or achieves something spectacular?

Like pay taxes?

Reply on news.ycombinator.com