Why I’m Calling BS on the Filter Bubble

The Filter Bubble is a fancy new term for “The Echo Chamber”, and although the idea is not new, we’ll hear a lot more about it in the next few years.

It’s always good to approach new technologies with caution and skepticism, but I am not particularly fussed about “The Filter Bubble”, certainly not anymore than I was ten years ago. Why?

  • New media actually has the complete opposite effect of a filter bubble. Instead of being forced to read all the views from a small handful of newspapers and TV stations, you have unlimited choice if you want it.
  • Personal recommendation systems, based on your browsing habits, don’t have to recommend things similar to what you’ve seen in the past; they can equally introduce randomness and are well placed to bubbling up opposing views.
  • So the choice is better than before, and so are the technologies to specifically break you out of the echo chamber. Then why do people end up in a filter bubble? Because they choose to. Most people are too busy with work and life to go out of their way to explore different news and views.
  • But you’re different. You do want to bust the bubble and you’ll specifically subscribe to challenging RSS feeds and follow Twitter users you disagree with. This is an awesome explanation of what probably happens if you attempt it. TL;DR: empirical evidence people queue up intellectual movies to watch later, but when push comes to shove, they will watch popcorn movies instead. We can assume the same happens when we sign up for different stuff. On the whole, the disruption hurts our head and we end up ignoring it, so the “filter bubble” is just a mirror of our own preferences.

There’s a good discussion along these lines in this week’s On the Media podcast (from around 33:00).

Watch It Later: A Comprehensive Review of four “Instapaper for Video” apps

In the past, I’ve tried pushing videos to Delicious under a “watchlater” tag or adding them to Instapaper, and neither of these options is effective. So it’s great to see this space coming alive, and how! Here are the options.


VHX is feature-rich, with hipster black background and stretching-to-the-edges full screen app experience. Login: Supports login via Twitter and Facebook. (I wish all the other services here also acknowledged site-specific login is passé. Lazy Reg ftw.)

  • Viewing experience: Uses an elastic layout where you can easily drag the side menu in and out to switch to a “almost fullscreen” view and it’s also easy to switch to completely fullscreen.
  • Discovery: The Channel feature is great – connects to your Twitter, Facebook, etc, as well as general media like BoingBoing, to find videos. There’s also a Dashboard to see your friends’ videos (and apparently the general user community if you don’t yet have any friends!).
  • Social features: You can “like” internal to the site and also share to social networks. The bookmarklet is also social.
  • Archiving: A “history” feature includes all the videos you’ve watched. (It was broken for me just now though.)
  • Bookmarklet and Adding: The bookmarklet identifies all videos and lets you click to add to your queue or share it immediately.
  • Other: One major annoyance is it launches straight into a video from the Dashboard, and this is exacerbated by the fact many of these videos are of the “funny viral” variety (Look! It’s a dog spinning around in circles. ROFLMAOCOPTER.) It would be a tiny change, but make much more sense to launch into your queue.


SQURL is also very feature-rich, with more of a “clean website” UI.

  • Login: No Twitter/etc here…need to make a new username/password, etc, you know the drill.
  • Viewing experience: Reminicent of YouTube in HD mode – video takes up most of the screen, with white margins. Controls for sharing/liking available.
  • Discovery: Discovery happens by curation – any user can create a new “gallery” and curate it, and it’s easy to browse these galleries. (Although adding a new video from here to your queue takes 4 clicks – 3 to many!).
  • Social features: As with SQURL, you can “like” internal to the site and also share to social networks. The bookmarklet is also social.
  • Archiving: All your significant actions, not just past videos, are shown in an Activities section.
  • Bookmarklet and Adding: The bookmarklet identifies all videos and lets you click to add to your queue. There’s also a Chrome extension to automatically stick Youtubes/Vimeos into your history.
  • Other: A major selling point is compatibility with many different video platforms, whereas the others seem to be about YouTube and Vimeo.


Instafilm is more about simplicity and has a neat slider to let you filter by time. (I wish it let you choose minimum time, not just maximum; on a day off, I sometimes want to watch a long video I’ve saved up.)

  • Login: Yep, it’s another startup who probably loses 50%+ of potential users because there’s no federated login support. Lazy Registration people!
  • Viewing experience: Didn’t like it as much as the others, found the video relatively small. I like the max-out approach best.
  • Discovery: No discovery. This is arguably a feature – the simplicity helps it to differentiate.
  • Social features: As with SQURL, you can “like” internal to the site and also share to social networks. The bookmarklet is also social.
  • Archiving: Hmm it lets you “Archive” videos, but this seems to mean “Delete”!!! I can’t find the video I archived anymore.
  • Bookmarklet: The bookmarklet is standard, but I did find it reports ok (“filmed”) whether it captures a video or not.


Like Instafilm, Radbox is a more down-to-earth experience that will be familiar to Instapaper users.

  • Login:Hurrah! It’s not just “create a new username and password”. You can at least login with Facebook, though other options would be nice too :).
  • Viewing experience: As with SQURL, a YouTube-like experience – video mostly maxed out on white background, with controls around it.
  • Discovery: No discovery. This is arguably a feature – the simplicity helps it to differentiate.
  • Social features: As with SQURL, you can “like” internal to the site and also share to social networks. The bookmarklet is also social.
  • Archiving: No archive as such, but you can add into a “Lists” as a form of categorisation. This feature, albeit basic for now, could make the service grow into a video an organising tool, not just “Watch Later”.
  • Bookmarklet and Adding: The bookmarklet is standard, and there are several other InstaPaper like tools available: a secret email address for inbound mail additions; an RSS feed; adding from Google Reader; and an API.


I’m on the fence between SQURL and VHX.

Key advantages of SQURL:

  • Platform support is a big deal here, and it seems to integrate with many more video platforms.
  • There’s already an iOS app (Android please). To be fair, I haven’t evaluated any of these on mobile, and SQURL does claim an optimised iOS web UI, though they’re also working on an app.
  • The curated galleries feature is excellent, they’re seeded it with a ton of galleries to begin with, and you won’t run out of interesting content anytime soon.

Key advantages of VHX:

  • Better UI. I like the video watching interface better overall, with its maxed-out videos, although some will find the initial jump into the Dashboard a little painful, and won’t stick around long enough to work out how the queue works.
  • One-click adding videos to queue, from within the app.
  • Exquisite bookmarklet. The way you can choose between adding to queue or sharing it immediately is awesome.
  • Channels are an excellent source of discovery, as an alternative to SQURL’s curated galleries, though so far, they weren’t niche enough to interest me. If they can build better curation into channels, ie custom channels made by users, that will be the best of both worlds.
  • Login via Twitter and Facebook.

For now, I intend to use VHX for day-to-day adding to queue…while using SQURL for its collections of videos in the event I feel like watching a random video that’s not in my queue. This may change fast though as it’s clear there will be a ton more innovation down the line.

I hope you found this useful, and please add any experiences or additional services in the comments.

The New Domain Game (aka “The Incredible Suckage of International TLDs”)

“.com” is over. Done. Late. Ceased to exist.

You will be incredibly lucky to find a green dotcom par excellence. All the good ones, the not-so-good ones, and even the outright mediocre ones are taken. So you have two options: buy an existing domain name or venture beyond the dotcom garden into the international TLD jungle. Buying an existing domain name has many of its own challenges, even beyond the sheer cost of the operation. But I’ll leave that aside for now as I want to comment on the fun of international TLDs here.

Let’s get something out of the way first. The TLDs are supposed to represent countries, I know. TV is Tuvalu, ME is Micronesia, etc. Well, sorry purists, but this is 2011 and .com is scarce. In the real world, I opt for cognitive ease over standards worship. The trend started with delicio.us and blo.gs, got steadily bigger over time, and will continue to do so at least until custom TLDs really take off.

The Minefield of International TLDs

Why are international TLDs so tricky to navigate?

  • Each registrar has its own set of TLDs it covers. Due to the bureacracy involved, they will usually just support a smattering of them, so if you’re holding a few names, you better start getting used to maintaining several registrar accounts. Ugh.
  • Some domains seem to only be handled by the original NIC, e.g. .io, so it won’t show up on meta-searches at big registrars.
  • Some domains have a policy of throttling requests at insanely frustrating rates. e.g. .io only seems to allow only about 8 a minute. .gy is similar. I appreciate they don’t want to get hammered, and especially by automated systems, but I don’t know how even legitimate users are supposed to find an optimal domain at that rate. At least offer more lookups via captcha!
  • It’s not clear which registrar manages which TLDs. Their own homepages do little to elucidate on this, often requiring you to perform a domain search before you find out. And even then, their interfaces don’t always make it clear.
  • When registrars do include info on their TLDs, they often break it down geographically – e.g. “European domains”, “African domains” etc.. That’s really irrelevant for most people’s needs. Alphabetical order would make far more sense.
  • There are different rules for each TLD. Some have length restrictions (e.g. .ly apparently no longer allows non-Libyans to register <4 letters). Some require a registered company in the place, etc. Some registrars will accept your order and then complain later (Gandi). Some will silently fail to search for the domain you wanted, instead suggesting other TLDs, without informing you why it didn’t do the search (Dotster).
  • Because of these problems, it’s apparently quite easy to get a high ranking for something like “.mn domain”, so SEOs have come along and made the search problem even more of a challenge. There are sites claiming to let you register a name, but which aren’t actually a registrar. At best, they probably just send your request on to the actual registrar, for a very hefty premium.

A Few Tips

On the whole, .com is still better. If you explain to a non-techie “my website is delicious”, what they’ll hear is “delicious.com”. Even if you say “that’s delicious as in del dot icio dot us”. Even if it’s a simpler name than that. People just equate web addresses with “.com”. this is even true with “.org”, many will still remember it as “.com”. Those perceptions will inevitably change, but not for another decade, which is about how long it took for naked domains – those lacking “www” – to sink in.

Domainr is a good resource. Not so much for the name suggestions, but the list of registrars against each TLD.

Gandi seems to have the most coverage. On the other hand, I’ve had weird experiences with them in the past, which is why I moved away many years ago. Now I just tried registering a domain yesterday, first time I used Gandi in 8 or so years, and it’s stuck listed as “waiting” with no explanation.

Try using whois or direct nic lookups (e.g. nic.io) rather than relying on the registrars, as you’ll find more TLDs.

As always with domains, be cautious about making queries. There’s always the possibility a site operator will yoink it before you get a chance. So “whois” is sometimes preferable or “dig” for the truly paranoid. And if you find a domain that suits, and it’s likely to be desirable, be ready to pounce.