The Morning Paper: a short summary every weekday of an important, influential, topical or otherwise interesting paper in the field of computer science.

There are several ways to take your morning paper:

During the course of a year, we’ll cover a broad range of computer science topics. I’m an expert in none of them! To the extent that I manage to bring any insight to the papers that I write up, this mostly comes from (a) reading a lot of papers across the various sub-disciplines, and (b) getting to see a lot of companies through my role as a Venture Partner with Accel in London. (It probably comes as no surprise if I also tell you that I spent the first 20+ years of my career in technical roles before finding myself at a VC firm!).

I’d like to take just a few moments to share with you what to expect from The Morning Paper, and offer a few suggestions for how you can make the most of it.

In terms of paper selection, my only real rule is that I find it interesting and that I believe it’s worth sharing. I’m naturally curious and love learning, so the set of things that I find interesting is pretty broad!  To read, understand (to the best of my ability) and write-up each paper takes me a minimum of two hours,  more often two-and-a-half to three hours and occasionally longer. My goal is to produce a piece that you can read and digest in around 10 minutes or less that gives you the essential ideas. I.e., something you can consume during a coffee (or tea!) break. Over the course of a year, you’ll be exposed to the ideas from 200+ papers.

I recently came across the following quote from Andrew Ng:

When I talk to researchers, when I talk to people wanting to engage in entrepreneurship, I tell them that if you read research papers consistently, if you seriously study half a dozen papers a week and you do that for two years, after those two years you will have learned a lot. This is a fantastic investment in your own long term development. (Inside The Mind That Built Google Brain: On Life, Creativity, And Failure)

Not all of us can make the time to seriously study six papers a week, but if you can manage to take on board the key ideas from five that’s still pretty good in my book! Please don’t feel pressure to read every single edition of The Morning Paper though. It’s fine to just read the ones that pique your interest, but I do strongly encourage you to dip into some of the papers from outside of your own area of expertise.

I believe we get the most out of a paper (and the ideas within it), when we actively engage with the material. Creating the write-ups is one of the ways that I do that, and having a few thousand people expecting their daily paper write-up to arrive in their inbox keeps me going on days when I don’t otherwise feel like doing it. (Thankfully, there aren’t many days like that – I love learning and sharing, I wouldn’t be able to keep going if I didn’t!).

Here a few suggestions for time-efficient ways of engaging with The Morning Paper write-ups that will help you get even more out of them:

  • Try explaining to a colleague or friend during a coffee or lunch break what today’s paper (write-up)  was all about.
  • Encourage a group of your friends and/or co-workers to also subscribe. (Just send them to this link). Get in the habit of chatting about the paper of the day socially in the office (or the chat forum!). Is there an analogy to some situation in your own environment?
  • You might even go so far as to create a mini reading group / discussion group within your company that meets once a week to discuss the week’s paper write-ups. It’s a bit like an in-house papers-we-love meetup, but without the difficulty of having to find speakers to present every time. If you do set up one of these, please do let me know. I’d love to hear any stories about how you’ve found it once it’s been up and running for a while too.
  • Forward a paper write-up to people you think might enjoy it and add a few comments of your own by way of introduction.
  • Share a paper that you enjoyed with your social network. I’m @adriancolyer on twitter, and I tweet each day’s papers as the write-ups go online.
  • Start a discussion on HackerNews or Reddit about a post that catches your interest. (I tend to focus just on writing the content – that takes enough time as it is! – but if readers help spread the word about the research I’m covering then that is always appreciated!).

If you’re a battle-hardened practitioner, I hope you find it stimulating to (re?)-engage with research ideas. With the gap from research to practice narrowing considerably in key areas of our field there’s plenty you can use, and I always like to cover papers that I believe practitioners can get immediate benefit from.

If you’re a researcher, I encourage you to especially take the ten minutes a day to read the paper write-ups from areas outside of your core research area. I believe many of the most interesting ideas and breakthroughs come from cross-fertilisation across the different (sub-) disciplines. Distributed systems, database systems, and machine learning communities all have plenty to teach each other as a simple example. If the paper of the day does happen to be in your area of expertise and you can add additional insights, please do leave a comment on the blog so that everyone can benefit. (Comments are moderated because spam, so sometimes it can take a while for them to appear).

If you’re an under-graduate or recent graduate, welcome! If only I had started this habit when I first graduated – think of all the compounding of ideas over the years!!

For whatever reason brings you to The Morning Paper, I hope you find at least one or two ideas every now and then that excite and inspire you, or that you can put into use. If that happens, I’d love to hear about it.

All the very best, Adrian.

About me: I’m a Venture Partner with Accel in London, where it’s my job to help find and build great technology companies across Europe and Israel.  If you’re working on an interesting technology-related business I’d love to hear from you: you can reach me at acolyer at accel dot com. Prior to joining Accel I spent over twenty years in technical roles, including CTO roles at Pivotal, VMware, and SpringSource.

Thanks for reading The Morning Paper!

79 thoughts on “About

  1. I found this blog today, searching for the title of one of my papers. I feel honored that you talked about it. And the blog went directly to my RSS reader, when I found some fantastic references to read. Great job.

    1. Hi Lu, thanks! I try to avoid papers that are heavily mathematical in order to keep the morning paper easily digestible. That still leaves plenty of scope for security-related material though. I’ll see if we can schedule something in for the 2nd half of the year – I don’t know the body of work in this area very well at all, but that’s a good reason to start reading!

      Regards, Adrian.

  2. I really enjoy reading the papers you find. Where and how to do you find new papers to summarize? I’m looking for more from the ICSE, but their website only seems to list titles of papers w/out any hyperlinks. Have any resources you can share?

    1. Hi Gabriel,
      There’s a paragraph in the middle of this post that describes some of the ways I source papers: https://blog.acolyer.org/2015/01/07/reflections-on-100-editions-of-themorningpaper/.

      I agree with you though that the ICSE website this year is very unhelpful! ICSE is on my list of conferences to keep an eye on, so I persevered, but couldn’t find any better process than looking to see what titles sounded interesting, googling for pdf copies of those papers, and whittling things down from there. It would have been much more useful to have had at least the abstracts on the website – often I’ve found great papers hiding behind innocuous looking titles… I downloaded maybe 15 papers in order to find the 5 I eventually selected for last week.

      Regards, Adrian.

  3. Very good! I am now looking into some interesting projects on graph computing and other similar big data applications, and I have a team working on GPU and FPGA accelerators.
    Stay tuned!

  4. Hi Adrian, this is great blog and I’ve been following for a month or so now, thanks!

    I was wondering what hardware/software you use to create the freehand diagrams and notes that often accompany your summaries? Are they drawn using a graphics tablet?

    1. Hi Alistair,
      Many of them are just drawn with pen and paper and then scanned. More recently I’ve been experimenting with Notability (and sometimes Paper) using the iPad Pro and the Apple Pencil…

      Regards, Adrian.

  5. Hi Adrian,

    I follow your paper reviews every morning…. Just too good.

    My question: How much time do you spend per review and writeup. As a graduate student I find myself struggle reading one paper a day, and writing a review is currently not even on the priority agenda. So what are the techniques that you follow that will help fellow grad students like me to improve reading/writing skills.. A blog post just on this aspect will be truly beneficial for all.

    1. Hi ‘testuser’,
      Many thanks for your kind words about the blog! I wrote a little about the process I use to read papers, and how long it takes (about 2-4 hours per paper, typically 2.5) here: https://blog.acolyer.org/2015/03/30/end-of-term-and-how-i-read-a-paper/ . Everyone is different, but for me the act of writing up notes on the paper really helps – otherwise I find I my retention of what I’ve read is not that great. And even with the notes, I often have to go back and re-read them (at least that’s quicker than re-reading the whole paper again!) when I’m thinking about a subject area.

      The only other thing I will say is that it does get easier with practice, both (a) you get better at quickly extracting and synthesising the key information in a paper, and (b) the more you read, the more of the background you already have which helps a lot too.

      All the best, Adrian.

  6. Dear Adrian, I just stumbled upon your blog via a retweet by raspberry pi. I subscribed immediately. Great blog! Thank you for the effort.

  7. Dear Adrian, it has already been some time since I started reading your blog. It is now becoming a habit of mine to check themorningpaper every time I get to the office in the morning while having my coffee. Your blog ranks along with Scott Aaronson’s blog, Shtetl-Optimized, in my list of personal favorites. I admire the discipline you have developed through reading and understanding papers with varying range of content complexity, almost every single day. I have not gone through all your previous reviews though but one thing I’d like to ask is, if possible, could you write a post about how you review papers? Do you have an outline maybe? I suppose after reviewing some considerable amount of paper you have some how come up with a personal guideline on how to write an effective review.

    Thanks Adrian.

  8. Hi Adrian, I found your blog a couple days ago. Soon it became my daily reading together with HN and etc. What I really admire is your discipline (as mentioned in the previous comment). At first I couldn’t believe there could be someone keep writing reports everyday, but you truly did. My respect to you, man!

  9. Hi Adrian, I really enjoy your blog. Please keep it up! Given your interests in databases I wonder what you think about the work of Alvin Cheung et al on combining programming language techniques and databases?

      1. Hm I’m not sure. Maybe their work on converting ordinary java code to spark? (along with their earlier work on compiling fortran to halide using similar techniques)

    1. I’m glad you like the blog, thank you. It’s fine to reblog or extract selected pieces so long as there is always a reference and prominent link back to the original source (e.g. ‘reposted from ….’ at the top of the piece). Thanks, A.

  10. Hate to sound like I’m repeating what everyone else says, but I’m not sure what else to say. Great work! I’ve signed up for the email.

  11. I am studying Computer Science, and your morning paper is really interesting to get track of what is happening. I wonder, how do you get your access to all the papers, or how do you get hold of the most latest one? Can you elaborate on it, so that I can get more on the papers? Do you have a curated list of journals or websites (arXiv, etc)?

    1. Hi Rakesh, I have indeed come to establish a curated list of conferences over time, there’s a mostly complete list here: https://blog.acolyer.org/2016/12/29/my-new-years-resolution-read-a-research-paper-every-weekday/. I also wrote a little about how I source papers in this post https://blog.acolyer.org/2015/01/07/reflections-on-100-editions-of-themorningpaper/ (see the paragraph starting “A good way to figure out…”).

      All the best, Adrian.

    1. Hi Dean, please feel free to quote as you wish! I missed Cosette first time around – the way they combine constraint solving and proof searching is neat and the results from the evaluation are impressive. Might have to put this on The Morning Paper backlog ;). Thanks, A.

  12. Hi Adrian, here another “thank you” from me! I really enjoy reading your write-ups every morning. I’m impressed how you manage to pick the interesting and important papers from such a wide variety of CS fields. And also that you write it up in such a way that my brain is able to follow early in the morning ;) Keep it up!

  13. Hi Adrian

    I discovered the ‘The Morning Paper’ about a year back and have been an avid reader since. I have tried to be regular in reading them each day and if at all I have to skip reading it, ensure that I makeup during the weekends. Thanks for making me a more better CS student and a professional.

    1. Hi Navin, that’s great dedication! I hope you’ve found the exercise valuable – I know it pays back for me in all sorts of ways I never could have predicted. Anyway, thanks for letting me know, it’s always encouraging to hear from people that integrate the blog into their daily/weekly routines :). Regards, A.

  14. Hi Adrian,
    How tough was it to transition from tech. industry to reading papers full of mathematical equations and CS-theoretical concepts? Did you have to revise some of the Maths and CS concepts?
    Have you been reading quite a lot already, before you started reading research papers almost daily?

    I think people loose the grip over CS theory once they join the industry. So, what helped the transition in your case?


  15. Hi Adrian, I really enjoyed looking through your blog! I was curious how you create your custom diagrams (what software, hardware?) — I would love to be able to do something similar for my own notes. Thank you!

  16. I hope you don’t mind one small bit of feedback about the email version of The Morning Paper (I would have emailed you, but it seems you do not want comments that way). The email edition currently does not come with a link to the paper under discussion — or even a mention of the paper’s title or authors!

    1. Hi Johathan, that’s a bug in my process! The email edition is most certainly supposed to include the citation and link. I’ll find out what’s gone wrong and get it fixed! Thanks, A.

  17. Hi,
    Take the liberty to bother you.
    We are volunteers of translation group from china,we found articles in “blog.acolyer.org” are very good. It’s very suit to developper. In China not all it developpers familiar with english. We want to translate you articles into chinese to make more chiese developpers and linuxers read it conveniencly for many chinese are not very familiar with english. We will keep author and website information in translations. We will publish the articles in our website : linux.cn. Can we get your license?

    Thanks a lot.

    Ezio, from Linux.cn

  18. Huge THANK YOU for keeping this blog going!

    I somehow stopped receiving The Morning Paper in late Jan, and thought at first you took a vacation. Then, realizing that I haven’t seen one in months, I got worried that perhaps you stopped writing altogether or (gasp) died, which would both be a tragedy. I’m so glad that neither is the case! Resubscribed and back to my daily dose of intelligence.

    I occasionally link to your posts in Tester’s Digest newsletter to spread knowledge in the testing community – thanks again!

    1. That’s super-interesting, thank you! I’d hypothesised you could use a discriminator network to reveal the presence of bias, but not seen a work actually doing it until now.

  19. Hi Adrian!

    You’re an inspiration :) Your commitment is laudable.

    I am wanting to start a blog on similar lines — easy digestable research.

    To follow the latest research, it’d be helpful if you could you list down some of the important research paper conferences you’ve come across/followed particularly in the area of distributed systems? For example I’ve found USENIX where Raft was published, which really interested me.


  20. Hi Adrian, I can’t believe that nobody told you yet that your blog is not mobile friendly.
    Great great contents, you succeed to find a format that is not primitive like usual tech news sites, but supplies a serious stuff.
    However, Morning Paper mails pile in my inbox, for a reason they have too small font to read on a mobile phone (I have an Android phone), and not being able to zoom in. The same problem when I try to read it on the site. The alternative is to read on a laptop (still the font is somewhat smaller than optimal for me). But for me getting to the laptop means starting a day’s work, so an article can wait.
    So it stops being a “Morning” paper for me. I still keep all unread articles, hoping one day to get to them, because it’s all really interesting stuff.
    Please consider to have a responsive format. It’s not even a novel trend, at least few years old. I believe it should not be a hard change to your blog template.

    1. Hi Michael, the site did have a mobile specific view that I see consistently from my own devices, but I don’t have an Android. I tried out the site with a variety of responsive design test tools though, and it didn’t do too well (not sure why the trigger to switch to the mobile layout didn’t fire there, but it didn’t). Readability is just about the only thing that matters on this site, so I’ve just switched to an alternative theme (and darkened the fonts for better contrast) that seems to do much better on e.g. https://responsivedesignchecker.com/checker.php?url=blog.acolyer.org&width=1400&height=700. The email issue will be separate because that’s a set of MailChimp defaults. The email is fully zoomable as-is on my iOS phone, I’ll have to look into what controls I have to see if I can make it better in other mail clients…
      Regards, A.

  21. Thanks so much for your work, I’ve really benefitted from your blog. Do you happen to know of any similar blogs which review Data Science research papers?

  22. Hi Adrian. I love your newsletter/blog! I have found a minor bug, and wanted to report it to you. But I can’t find a “feedback” type of mechanism. Can you let me know what mechanism would work best to report it to you?


  23. Hi Adrian. I love your newsletter/blog! I have found a minor bug, and wanted to report it to you. But I can’t find a “feedback” type of mechanism. Can you let me know what mechanism would work best to report it to you?

  24. Thank you, Adrian, for keeping this astounding blog alive. I read as much as possible and only wish I could be as diligent as you are! It’s a pleasure to be part of your readership.

  25. Of possible interest:

    From: calendar@csail.mit.edu
    To: fast-code-seminar@lists.csail.mit.edu, seminars@csail.mit.edu,
    Subject: TALK: Monday 03-16-2020 How fast are Algorithms Improving?
    Content-Type: text/plain; charset=UTF-8

    How fast are Algorithms Improving?
    Seminar Series: Fast Code Seminar 2019
    Speaker: Neil Thompson
    Speaker Affiliation: MIT
    Host: Julian Shun
    Host Affiliation: CSAIL MIT

    Date: Monday, March 16, 2020
    Time: 2:00 PM to 3:00 PM

    Location: 32-D463 (Star)

    Abstract: Quantitative analyses of how fast algorithms are improving are usually done ad-hoc. Longer term analyses are quite rare – in fact, almost all use the single example of Linear Solvers. In this paper, we gather data on 127 different algorithm families from more than 40 textbooks and 1000 research papers to present the first comprehensive look at algorithm progress ever assembled. We find that as a whole algorithms have undergone substantial progress, but with huge variation: in some cases, progress rivals those from other important sources (e.g. Moore’s Law for computer hardware), but in other cases it shows relatively little progress.

    Bio: I am an Innovation Scholar at MIT’s Computer Science and Artificial Intelligence Lab and the Initiative on the Digital Economy.

    I am also an Associate Member of the Broad Institute. Previously, I was an Assistant Professor of Innovation and Strategy at the MIT Sloan School of Management, where I co-directed the Experimental Innovation Lab (X-Lab), and a Visiting Professor at the Laboratory for Innovation Science at Harvard. I have advised businesses and government on the future of Moore’s Law and have been on National Academies panels on transformational technologies and scientific reliability.

    I did my PhD in Business and Public Policy at Berkeley, where I also did Masters degrees in Computer Science and Statistics. I have a masters in Economics from the London School of Economics, and undergraduate degrees in Physics and International Development. Prior to academia, I worked at organizations such as Lawrence Livermore National Laboratories, Bain and Company, The United Nations, the World Bank, and the Canadian Parliament.

    Relevant URL: http://fast-code.csail.mit.edu/
    For more information please contact: Julian Shun, jshun@mit.edu

Leave a Reply to Rakesh Silwal Cancel reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.