Surfing Ephemera Overload

The information firehose we’re blasted with daily mocks us to focus. It consumes and divides our attention and defiantly laughs at our plans.

Pre-COVID, I outlined how I changed things up to deal with the fact I was grinding rather than progressing critical information consumption.

The reset definitely helped consumption-wise, but ultimately I want to be applying knowledge I gain. In particular, I was still having issue with what I describe as the ephemera class of information, so wanted to see if it I could improve my flow, and make sure I was extracting the important (and discarding the unimportant faster) to leverage longer term gains and opportunities.

So What’s the Goal?

To be clear on where we want to get to. I want a process to:

  1. Limit
  2. Capture and triage (while also not artificially narrowing sources, topics, or opinions)
  3. Sequence and consume
  4. Absorb, learn, and action

If you read the 2020 post will note that I’ve riffed slightly on the original 3 items to include one on sequencing and consuming, and another on how to leverage information. “Processing and work”, the original, vastly understated what was required in the flow.

To start off with, let’s go over what worked and didn’t.

So What Worked?

Certain themes worked across everything:

  1. Constraining to high quality sources
  2. Limiting Inboxes
  3. Deliberate, explicit consumption

1. Constraining to high quality sources

While we all need time-wasting crap to let our brains go slack once in a while, removing that as a default (and creating friction to have to get to it) and having subscriptions to high quality information sources (eg. NYTimes, Economist, courses, newsletters, and youtube channels) has worked wonders (for some reason, I just can’t do audio podcasts).

The hard part here is avoiding creating an “info bubble” to make sure you still get competing and contrary opinions so you aren’t in an echo chamber (harder than it sounds with news and social media algos these days). Bottom line: Great reporting and in-depth analysis from news to science to finance is in decline across the board and you need to search for quality (as well as the people continually noticing the important things others are not).

2. Limiting Inboxes

It’s insidious how much time we spend “processing” and reacting to inbound information (and often re-handling it). And how much pushed information are hooks to our attention and a major source of distraction and considered “work” (I have literally seen execs spend their entire days pushing emails and slack messages around).

I personally try to make sure I have a limited number of inboxes I check and process daily and a simple, solid process to get down to “inbox zero” in each, extract actions in each case to my GTD app and try to maximize time working on things rather than handling and shuffling information.

3. Deliberate, explicit consumption

As opposed to automatic or even unconscious, pushed “on auto” consumption (yeah, looking at you Netflix) meant setting routines and habits around when and how to look at things. I’ve even scheduled time to check and process (or do explicit review of material) and having defaults mostly around working through lists of stuff (to not make it sounds fancy). The process - extract - make list - work through list has definitely made me vastly more productive. I still backslide when I’m tired or fatigued, but much less so than before.

How Did Each Class of Information Do?

But, to analyze what didn’t work, I’m going to split information into three classes since they had different dynamics and successes, and I feel only one needs major overhaul. They’re:

  1. Media
  2. Learning, and
  3. Ephemera


Books, movies, series, and even games ended up much better handled over the last two years (less so games which finding time to play through and examine mechanics I did less well at).

The key came down to two things:

  1. Have an explicit ordered list

    For each category of things I wanted to read, watch, or play and deliberately working through them one at a time in order (or explicitly re-ordering and nothing why to myself).

    Another hack here was to fix the list in terms of what I could accomplish in a year realistically and then any additional things (esp books) I wanted to read were appended to an additional list and I would only introduce or remove things from the list deliberately once a month (ok, sometimes I’d just insert a book I particularly wanted to read - or to avoid a book I didn’t).

    Also, timeboxing and scheduling a routine for things like reading made a big difference, and I even went one further than turning off autoplay on services like Netflix after covid and unsubscribed from Netflix and others entirely (now I just one-off series or movies I really want to see which feels like it has freed up enormous amounts of time. Streaming services are scary time sucks due to the low friction of continuing watching. Now, if I wanna binge watch something, I explicitly schedule it - usually with a friend and pizza.).

  2. Creating a Resonance Calendar

    Where I take notes on individual items (esp books) and then doing a review and creating key action items… as well as notes, questions, and prompts for further study or followup to make sure I understand key concepts in books well. This is an idea I shamelessly stole from Ali Abdeel but which has been pure gold in terms of helping to leverage and bake-in learnings, particularly from non-fiction books. I still need some work here for this to work as well for textbooks or coursework which is harder to “summarize”. And also, while his technique works well for Notion, it was harder to apply to other GTD apps like emacs and my current GTD app logseq (mostly due to the lack of a good capture mechanism from web and mobile for both. Notion’s capture plugins for browser and mobile are excellent.). The astute amongst you may note this is simply another list, but it’s worked very well.

    Surprisingly, Kindle highlighting and services like Readwise have been much less useful than I would have hoped and it seems a good end of book (or chapter) review, solid notes and summaries in my own words, and action todos to leverage what I’ve learned or expand on the foundation has been the most useful. As an experiment, I’m killing Readwise and Instapaper for the near term to see if having a more explicit review process works better (I do like how Readwise surfaces old quotes from old books in my mail though, but perhaps I should be replacing that with some sort of “random book I’ve read” review more explcitly.


Primarily coursework and textbooks, I’d say this has gone very well in learning hard science, primarily astrophysics and bioengineering (more astro in the last year), but less well in learning new programming languages and IT concepts.

Much like the media section above, generating a course of study (ie. a list) to figure out what I want to learn, and then working through that as a (so far, neverending) work in progress every week has worked best (though at times could feel slow and frustrating depending on the quality of the course material. Particularly for online courses, the quality of some vary wildly as does the amount of mule work in some.).

In coding, I feel most of the problems are more me falling prey to the technical press who keeps on putting shiny, new languages and techs in front of you, so that I spread myself too thinly between multiple coding languages rather than doubling down on one for mastery (I shemefully hopped between Go, Rust, and Typescript the past year though feel I’ve now just decided to double down on Go, warts and all). Hopefully I’m over that now and just focusing on mastery.

Ephemera, News, and Crow’s Nesting

This is the real area I tended to struggle with goal-wise (and easy to dismiss lack of traction, since it’s often aspirational and pile-on.)

Large amounts of newsletters, updates, events, and opinions that cascade in via email, web browsing, blog posts, newsfeeds, friends, or twitter and where I collect, then process, then try to extract what is useful and find time to actually action and deep dive into the material that I consume there (I’m also going to add research papers into this now since that’s become a thing the last half year.).

In particular, the sheer volume of material that comes through often means I had a tendency to save things to a Read Later app (Instapaper) but found that it ended up being a junk drawer rather than Inbox and I spent too much time processing (possibly due to its interface which is optimizing for reading rather than processing) or ignoring it growing out of manageable proportion. Unlike say, my newsreader (newsboat) which let’s me rip through large volumes of info like a chainsaw, I spent too much time adding to the lists rather than processing and extracting from them to work on.

As an example, ahead of this post, I exported my Instapaper and found there were over 7k (!!!) items I had “Read Later”-ed but when I culled through that export, stuff I actually wanted to Read and process came down to about 500 (so yes, the astuate amongst you probably realize this means I declared inbbox bankruptcy on Read Laters and am changing things up. Thus the post.)

Needed a better system to triage, and also one that distinguishes between reading later and bookmarking as references for future referral.

Perhaps also, a dose of realism myself in terms of how many items I can add to a Read Later bucket on a weekly basis that is not delusional (rather than aspirational) in what I can work through and investigate comfortably. And yes, partly this may have to do with more disciplined culling (and not saving for Read Later at all - much like everyone I think I need to be a bit more immune to link bait posts that are thinly veiled business self-help posts which are low value content mill work or simply paid shilling by companies.).

Characterize the main problem here, it’s the ability to frictionlessly bookmark Read Later material into a bucket to be dealt at an “elastic later” that doesn’t exist for me time-wise. I believe also constraining my list visually (to what I can see) so that i can assure I can then cull when item numbers get too big. Some items I should have extracted to my Someday/Maybe list in any case since they were techs/languages/techniques that were highly exploratory. Some of these fell naturally off the 7k list when I was cullin, being superseded or lacking traction.

How I’ll Change Up Ephemera

Strangely, I think part of the problem here may hve also been the tool if not the approach, though there are a couple of great things about my Instapaper approach to double down on. What I liked about Instapaper was its ability to forward it newsletters (particularly technical ones with lots of links) from an auto mail processing rule in my inbox. This kept my main inbox closer to inbox zero and moved ephemera into the “Read Later” pile. It also has a great, one click browser and mobile bookmarking extensions. Genius.

But, as mentioned, that created new problems:

  1. An infinitely long, “invisible” queue (not connected to time I have to absorb or work on items)
  2. Muddling of Read Later with Reference bookmarks (and no organizational hierarchy)
  3. Yet Another Tool and slower management and processing interface (compared to say ripping through things in the original mail newsletter or newsfeeds)

I think there are four tweaks to fix this:

  1. Separate Read Later from Saved bookmarks and use PARA to organize them hierarchically
  2. Create a Read Later label in my mail and re-code the rules to bypass the inbox and tag them
  3. Have a distinct Read Later tool and flow
  4. Schedule specific time for reading and working on Read Later material

1. Separate Read Later from Reference and PARA

Probably the first step is to winnow out things I want to read later from those that are save for some unforeseen future reference (house designs, for example, or something I’ve read which may have use long in the future.). Keep Read Laters separate.

I’m taking inspiration from one of my previous staff who had the most elaborate bookmarking hierarchy I’ve ever seen (for work). While slightly crazy (and hilariously German), it being highly organized and logical meant he found things impressively fast (my work has a problem with needing to create spreadsheets to track all the spreadsheets it creates, though not sure there is a better solution for such a sprawling global organization running mostly on google docs.).

So, the idea here is to reorganize what is Reference and use Tiago Forte’s excellent PARA approach (Projects, Areas, Resources, and Archives) organizing hierarchy and use synced in-browser bookmarking rather than a separate tool (The hierarchy will reflect the one in my GTD system.).

My hypothesis is knowing where to place stuff and Marie Kondo out of the Reading List will clearly delineate what has to be dealt with versus what is saved for contingency (though, would love it if in-browser bookmarking also saved the actual content of sites. I was amazed if the 7k how many had disappeared due to link rot.).

2. Read Later mail label and auto processing rules

So, one of the easier fixed was repointing Instapaper forwarded rules to a “Read Later” label in my Gmail.

Yes, I bypassed one inbox to create another, but that helped in two ways:

  1. I have a lower priority inbox which I can process in bulk and in the same tool
  2. I am way more efficient buzzsawing through mail so an implicit speed up over Instapaper

3. Specific tool and flow for Read Later

This may just be a matter of simple tools work best, but having one place I can see visually how many Read Later items I have instantly felt like a better approach. Perhaps it’s the ability to see everything at once, or the ease of deleting and reordering (often on later reflection, Read Later items seem not worth the time in light of a new day.).

So far, using One Tab for this which is a Firefox (and Chrome/Chromium) extension has been great. You can right click on a browser window and throw that tab into the one tab list (and create named lists). I keep it as a pinned tab at the far left of my browser. Trying to figure out what a good, realistic queue size (and what gets pushed off the table if I go over) is something I think I’ll need to experiment with.

4. Schedule specific Read Later review and work time

Strangely, considering how I feel I’m pretty effective at this in other areas, Read Later time always seemed to be optimistic and aspirational time. This was an error. Unlike better habits I now have with reading, writing, coding, and Astro, I never scheduled explicit time in for Read Later. Not only did it make me feel like I was failing but it also made me possibly not capitalize on new opportunities I should really be crow’s nesting for constantly. Limiting topics and timeboxing/scheduling has worked so well in other places it’s a bit ridiculous this was a blind spot for me.

I think a key problem I want to get out ahead of will be ephemera sparking other todos and areas of investigation which will themselves need extra time and sequencing. Time is our scarcest resource.


So, that’s the idea: Simpler tools, separation of concerns, a constrained list, and explicit time to align the firehose with my time available and be more effective. I’ll let people know how it’s going in a followup post in a few months’ time.

As always, I hope this post helps you be better and accomplish more of what you want to get done calmly and efficiently. I’m always curious to hear how it’s gone for people who may adopt some of these processes, or hear more about what may have worked (even better!) for you or other things that have made a huge difference for you. Feel free to mention me as @awws on mastodon or email me at via email


b3561cd @ 2023-01-19