Topic: Data Mining

Senator Marsha Blackburn raised alarms about Meta and Google allegedly coordinating to listen to phone conversations for keyword-based advertising. Blackburn argued that federal online privacy protection legislation has been repeatedly blocked by high-priced lobbyists. The hosts discuss how users unwittingly grant microphone and contact access to apps, making themselves the product.

The hosts express skepticism regarding the long-term viability of the current AI boom, comparing it to previous tech bubbles. They discuss the massive energy requirements for new data centers and speculate that if the AI industry collapses, the excess power infrastructure could be repurposed for Bitcoin mining.

The CDC admitted to providing false information regarding its analysis of vaccine safety data in the Vaccine Adverse Event Reporting System (VAERS). While the agency initially denied having reports on post-vaccine heart inflammation from before October 2021, it later acknowledged a connection existed as early as May 2021. Additionally, discrepancies were found in the timeline of when the agency began performing Proportional Reporting Ratio (PRR) data mining.

A discussion of historical software algorithms reveals how companies can mine internal emails to identify untapped expertise within their workforce. While such practices raise significant privacy concerns, the potential value of analyzing public data on platforms like Twitter is estimated to be worth billions. The hosts suggest that Twitter's previous management was too focused on political moderation to exploit these data-mining opportunities.

The Texas State Teachers Union and other labor groups are lobbying to delay school reopenings, while New Hampshire Governor Chris Sununu has pushed for in-person classes. Amidst the shift to remote learning, Google Classroom has grown from 15 million to 100 million students, raising concerns about the long-term data mining of children.

A listener anecdote illustrates how Google mines Gmail data to target advertisements, such as moving a company's emails from spam to the primary inbox based on search history. The hosts debate the reliability of decentralized email protocols versus centralized providers like Gmail and ProtonMail.

PBS Frontline launched a transparency project by releasing the full, unedited interview with Trump digital director Brad Parscale. The project aims to provide context on how the 2016 campaign utilized Facebook's data and algorithms. The move follows ongoing investigations into whether Facebook sold or shared user data with political consulting firms like Cambridge Analytica.

Updater.com is a service marketed to college students and millennials to help manage the logistics of moving. The hosts reveal that the company's business model involves selling "contextual and personalized" data to businesses trying to reach relocating consumers. They warn that users are "being taken" by providing credit card numbers and subscription details to a platform that will eventually be hacked.

The "Stylish" browser extension was found to be harvesting and uploading users' entire browsing histories. Additionally, reports reveal that third-party developers and Google employees have been able to read unredacted Gmail messages to train AI algorithms. The segment argues that the "AI" used for these services is often just a series of manual grinds and human oversight rather than true machine intelligence.

A 2012 presentation by Carol Davidson of Rentrak reveals that the Obama campaign "rocked" Facebook's privacy policies to ingest the entire social network of the United States. Unlike the 50 million users cited in the Cambridge Analytica scandal, the Obama campaign had access to the full network because Facebook had not yet restricted "friends of friends" data access. This "smoking gun" clip suggests a massive double standard in how the media treats Democratic versus Republican data mining.

Representative Maxine Waters previously praised President Obama's massive, unprecedented database of individual information, which was later transitioned into the 501c4 group Organizing for America. The segment contrasts the media's praise for Obama's data mining with the current condemnation of Trump's data usage. Additionally, they touch on emerging liberal narratives claiming the 2016 election was "stolen" or "hacked."

Austin voters rejected Proposition 1, leading Uber and Lyft to suspend operations within city limits due to mandatory fingerprint background checks. The discussion posits that the conflict is actually centered on ride-sharing companies' refusal to share granular passenger data with the City of Austin. Comparisons are made to data mining practices at American Express and the high valuation of user data over transportation revenue.

The MIT Technology Review published a study claiming that data mining Twitter can reveal the origins of support for the Islamic State. The hosts dismiss the study as a way to justify increased surveillance and social media monitoring. They play a song about the loss of human connection in the age of digital "likes" and "shares."

Major tech companies including IBM, GE, and Verizon are forming a consortium to create common standards for Hadoop, an open-source software used for large-scale data processing. While the industry promotes Hadoop as a revolutionary tool for extracting useful information from vast datasets, skeptics argue that "big data" is often a money-sucking scam that fails to deliver truly relevant consumer insights or effective targeted advertising.

The discussion explores why major tech companies like Google and Facebook have not simplified encryption for average users. It is argued that true privacy is antithetical to their advertising-based business models. Steven Levy's defense of these companies—claiming they don't "sell" data but are "reimbursed" for supplying it to the government—is dismissed as a semantic distraction from the reality of data mining.

The hosts criticize the "Sign in with Facebook" feature found on many websites and apps. They explain that this is a data-mining tool for developers to access a user's personal information, preferences, and friend lists. They argue that users are the product in this ecosystem, which is designed to satisfy venture capital investors rather than provide a free service.

The hosts reiterate their preference for open-source genetic analysis over commercial services that mine user data. They link the arrogance of 23andMe to broader tech industry trends, including Google's data practices and the monetization of NSA documents by figures like Glenn Greenwald.

Adobe's acquisition of Omniture for $3 billion is framed as a move into the government intelligence sector. An Adobe white paper titled "Adobe Insight" describes accelerating government intelligence by combining cell phone, GPS, and email records into actionable knowledge. Adobe's data center in Utah's "Silicon Slope" is noted for its proximity to the new NSA data center.

Google's public opposition to National Security Letters is analyzed as a business decision rather than a move to protect user privacy. The company already mines data for advertising and reportedly sells information to various international entities. The legal pushback may be an attempt to force the government to pay for data access rather than providing it for free under administrative subpoenas.

Financial institutions and companies like Federal Express use sophisticated data mining to profile consumer behavior and detect suspicious patterns. The US government is reportedly seeking access to all credit card transaction data to monitor political affiliations and spending habits. Specific purchase patterns, such as buying gas and Nikes in quick succession, can trigger automatic account freezes.