Alpha60 is a long-term study of media on the distributed internet, focusing on streaming film and television, but also including examples of other media including 3d gun files and hacks/leaks from the transparency collective Distributed Denial of Secrets.

This fall, we will start with groupings of streaming tv, and depending on the progress we make with those data sets, we will be widening the scope to include the UKR-RUS cyberwar leaks.

About our technical background and scope.

There are three parts to alpha60 compute infrastructure:

  1. Samplers: cluster of 9 leased bare metal alma 9/10 linux machines in europe that run a custom C++ p2p swarm sample application 24/7. This cluster is administered in python with ansible, reads a github data repository for input, and is scheduled via chron. Standard operating practice is to do super-saturated sampling of the p2p swarm every 3 minutes, for 10, 15, or 26 weeks. Each sample (aka every 3 minutes) writes a json output file. These files are archived and compressed, and transferred to offline storage.
  2. Caching: this is a cluster of workstation hardware in SF that aggregates individual samples into aggregates in durations of hour, day, week, and cumulative.
  3. Analysis: another cluster of hardware that takes aggregates, does IP->geolocation, and writes out an anonymized json file stripped of ip addresses but augmented with geolocation and network characteristics using specialized ip to geolocation software from Ipinfo.io. In fact, alpha60 has research access to all Ipinfo data APIs at the highest resolution in unlimited quantities (with attribution). This group will primarily be working on this part of the infrastructure.

Our scope:

Year a60 download netflix subscribers
2018 63M 139M
2019 154M 167M
2020 206M 204M
2021 594M 222M
2022 353M 231M
2023 743M 260M
2024 1.99B

We are going to be developing new techniques for data analysis of this information. This is new territory, and we will be trying to correlate to existing media metrics from film (box office) and streaming tv (Nielsen, and look at some of the other metrics on a per-platform basis.) We will be trying to estimate the growth of piracy vis-a-vis growth of the internet and growth of commercial streaming platforms Disney+, Amazon Prime, Netflix, Apple TV+