February 25, 2018

Telex and Ethan Zuckerman's "Cute Cat Theory" of Internet Censorship

A few years ago, Ethan Zuckerman gave a talk at CITP on his “cute cat theory” of internet censorship (see also NY Times article), which goes something like this:

Most internet users use the internet and social media tools for harmless activities, like looking at pictures of kittens online. However, an open social media site is open to political content as well as pictures of kittens. Repressive governments might attempt to block this political content by blocking access to, say, all of Blogspot or all of Twitter, but in doing so they also block people from looking at non-political content, like pictures of cute kittens. This both brings more attention to the political causes the government is trying to suppress through the Streisand effect, and can politicize users who previously just wanted unfettered access to cute kittens.

This is great for Web 2.0, and suggests that activists should host their blogs on sites where a lot of kittens would be taken down as collateral damage should they be blocked.

However, what happens when a government is perfectly willing to block all social media? What if a user wants to do more than produce political content on the web?

Telex (blog post) can be seen as a technological method of implementing the cute cat theory for the entire internet: the system allows a user to circumvent internet censorship by executing a secret knock on potentially any web site outside of the censor’s network. When any web site, no matter how innocuous or critical to business or political infrastructure, can be used for a political goal in this fashion, the censorship/anti-censorship cat-and-mouse game is elevated beyond single proxies and lists of blockable Tor nodes, and beyond kittens, to the entire internet.


  1. Alastair Carnegie says:

    One of these days, we are going to set the cat among the pigeons, and publish the factors of RSA1024 on Wikipedia…..For those who wish to factor it themselves, a clue was given several tears ago on TechRepublic by an alter-ego friend Brother Martin de Porres. “Integer Factorization and The RSA Problem” I presented a trivial worked example that could just fit the 32 digits available with windoze calculator. The protocol is based on basic grade school algebra. e.g. 10057 = 89 x 113, and 10201 = 101^2, and 10201 – 10057 = 144, and 144 = 12 x 12, and 101 + 12 = 113 and 101 – 12 = 89. The trick is to find an approximate ratio of the two composite primes. Initially by trial and error, say 2:3, or 3:5, or 4:7 whatever? then apply a Goldilocks ‘Squareness Test. we know that if P & Q are primes, and p & q are in the same ratio, then a rectangle Pq x Qp will be almost square. We can compare our test ratios against PQ^2 which we know is perfectly square, and assign a Figure of Merit “Z” which will converge ever closer to unity, with each Goldilocks estimate ratio….I was happy that L.Bontes, after tinkering with a few of his own randomly generated “PQ” composites, declared the method “Faultless”.