Rep. Thomas Massie: Government Collecting Millions of Exabytes of Data Without Warrant

This lecture is technical but, as it is a course-overview, non-techies will glean useful insights from it, so I recommend it to anyone who would like to understand the theoretical underpinnings that make modern mass-surveillance possible.

The key for non-technical people to understand is that the kinds of digital systems that have been deployed as part of the architecture of mass surveillance of US citizens are indefinitely scalable. That means that these systems maintain the same responsiveness to queries when they have a million bytes as when they have a million exabytes. Users of consumer computing equipment like iPhones, Androids, and Windows PCs are used to the "slow grinding degradation" of performance over time, but this is not the result of any inherent requirement that systems with more files/data/apps must become more slow over time, it's just crappy corporate software developers who don't give a damn (and don't have to give a damn, because you'll throw your device away after 5 years of use, anyway.) But the Five-Eyes surveillance systems that Snowden disclosed are designed from the ground-up for indefinite scalability. This means that, as technology is able to capture more and more detailed information about our daily lives, that data can be siphoned into the Big Brain at Bluffdale and it will be searchable for mass-surveillance purposes with exactly the same responsiveness. In 2005, your cellphone knew where you were at all times to within about 1 meter. By 2010, it knew within centimeters. Today, your phone can tell its position within a centimeter, it knows its elevation above sea-level to within a meter (it knows what floor you're on in a multi-story building), which way it is oriented in 3D space and, if a camera is on, it can map its environment in 3D, in real-time; it can transcribe every single word that is spoken in its hearing-range (which is better than a typical human's especially in noisy environments) and forward all that now "meta-data" back to the mothership at Bluffdale, to be stored forever, and data-mined indefinitely. Whereas your desktop PC bogs down the more files you load into it, the indefinitely-scalable military mass-surveillance system keeps its same frame-rate year after year, even as the resolution of the data it is harvesting is increasing exponentially. As developments like ChatGPT make clear to the public now, the old "information overload" theory of why mass-surveillance can't work doesn't apply to these systems. A text-summarization AI can summarize and compress harvested data with 95+% accuracy, allowing all conversations to be indexed not just by keywords but semantically, that is, in terms of the substance of the topic being discussed. And queries on these ocean-scale databases are effectively instantaneous, due to the built-in scaling that they were designed with.

Don't make the mistake of underestimating the enemy! And yes, the mass-surveillance state is the enemy.

Jump to timestamp 28:00 to skip the housekeeping...
 
Last edited:
Back
Top