Topic: Data

Marty and Dana Brunetti offer their parting thoughts, with Marty highlighting how Dvorak supported him through complex open-heart surgery. Dana reflects on Dvorak's ability to back up his strong opinions with history and data, making people think even when they disagreed. The segment ends with a humorous anecdote about Dvorak's frequent "butt dials" and a plan to meet in Oakland on August 8th.

An analysis of declassified documents referenced in Donald Trump's speech examines claims of Chinese interference and the acquisition of 220 million US voter files. While some documents suggest voter data was purchased from commercial sites rather than hacked, others detail FBI investigations into fraudulent voter registration in Michigan. Trump advocates for the passage of the SAVE Act to require proof of citizenship for voting and eliminate most mail-in ballots.

The Chinese Ministry of Foreign Affairs responded to Donald Trump's allegations, describing them as "pure fabrication" and "malicious smears." While reports acknowledge China has collected publicly available American voter data for years, officials deny this allows for vote manipulation. The discussion also touches on China's strategic interests in Iranian oil and the Strait of Hormuz.

Data centers operated by Meta, Google, and Amazon are criticized for their massive water consumption, with some using up to 30 million gallons to stay cool. Reports indicate that local wells are running dry as a result. Additionally, concerns are raised about wastewater contamination involving "forever chemicals" (PFAS) and heavy metals that disrupt local treatment facilities.

Protests occurred across the Bay Area, including a demonstration on a pedestrian bridge in Berkeley against AI data centers. Simultaneously, a national call to action against ICE took place in Redwood City following fatal shootings in Maine and Texas. The "Good Trouble" weekend of action, honoring the late John Lewis, also saw events focused on civil rights and voting rights.

Donald Trump is promoting a plan to fast-track approvals for energy plants, including small nuclear reactors, to power AI data centers. This initiative is framed as a potential "Trojan Horse" for building a distributed national energy grid that provides cheap electricity regardless of AI's ultimate success. Trump claims to have personally instructed tech leaders like Mark Zuckerberg and Jeff Bezos to build their own power plants to avoid straining the existing public grid.

Palantir has entered a direct partnership with Nvidia to allow customers to run open large language models while maintaining control over their own data and "alpha." Alex Karp argued that enterprises want to own their means of production rather than transferring data to third-party providers. He also dismissed the "safety" marketing used by AI companies as a gimmick to avoid government oversight.

Investor Kevin O'Leary is reportedly behind a massive 40,000-acre data center project in Utah, which is roughly twice the size of Manhattan. Critics and local activists have raised concerns about the project's massive electricity requirements and its potential for mass surveillance. The facility would be significantly larger than the existing NSA data center in the same state.

The hosts analyze the global rush to build "hyperscale" data centers as a potential real estate scam or investment bubble. They argue that the physical size of these proposed facilities, such as the Utah project, makes little sense for actual compute needs. The discussion also addresses public fears that these centers are for mass surveillance, noting that surveillance is already primarily conducted via smartphones.

The hosts advocate for local computing and the use of open-source operating systems like Linux to avoid data harvesting by large corporations. They criticize the public's willingness to upload personal data to the cloud while simultaneously fearing government surveillance. The discussion emphasizes that privacy is a choice made through hardware and software habits.

A technical error occurred when an AI-driven "Stage" skill used an "rsync --delete" command, resulting in the accidental deletion of several critical production folders, including donations and show notes. The AI rationalized the deletion because the folders were not present in the local mirror it was instructed to sync. This serves as a cautionary tale about the dangers of allowing AI to run autonomous processes without robust backups.

Palantir CEO Alex Karp expressed skepticism regarding the ability of large language models from companies like Anthropic to replicate Palantir's enterprise-grade engineering. Karp criticized the culture of Silicon Valley and warned that a failure in the current AI hype cycle could trigger a massive market correction for tech stocks.

Former Pfizer researcher Michael Yeadon claims that the COVID-19 pandemic was a fraudulent event exaggerated by inappropriate PCR testing. He argues that all-cause mortality did not increase significantly prior to the WHO declaration and that many deaths were caused by changed medical procedures, such as inappropriate ventilation and the denial of antibiotics.

Due to the high cost of cloud-based AI, Fortune 500 companies are beginning to shift toward running smaller AI models locally on "desk-side" hardware like the Mac Mini. This transition creates new challenges for network bandwidth and "telemetry events" as local agents coordinate with cloud resources, providing a new sales opportunity for networking hardware companies.

Financial analysts report that data center projects are now requiring $30 billion to $50 billion in debt financing, a staggering amount for the loan markets. On CNBC, a proposal is made for a "modest AI token tax" to prevent massive unemployment caused by automation. The segment also highlights concerns that AI companies are using $100 million Super PACs to silence political representatives and prevent regulation.

A proposed bill in Minnesota seeks to limit the driving of classic cars to weekends and special events, sparking concerns among enthusiasts. While the bill may be tied to cheaper registration fees, the hosts speculate it is a move to phase out older vehicles that lack modern tracking technology. They suggest this fits into a broader narrative of increasing surveillance through data centers and license plate readers.

A report from BBC Future reveals that modern cars collect vast amounts of personal data, including weight, facial expressions, and destinations. John Dvorak advocates for driving classic cars to avoid this surveillance. The hosts then credit "Blue Acorn" for the Mother's Day-themed artwork for episode 1867, which featured a "Mama Hanta" rat surrounded by virus molecules.

An anecdotal report from a Google finance employee suggests that the Gemini AI group is profitable due to total supply chain integration. Unlike competitors, Google utilizes its own data centers and proprietary chips rather than renting infrastructure or buying from Nvidia. This vertical integration, combined with the vast data from YouTube, provides a significant market advantage.

Ballistic missile and drone strikes have targeted AWS data center locations in Bahrain and the UAE, causing devastating physical damage. The vulnerability of expensive AI infrastructure to direct military hits is highlighted as a new risk in modern warfare. Some suggest the conflict is tied to a broader strategy to ensure U.S. dominance in AI by controlling the energy required to power massive data centers.

Donors contributing $200 or more are granted "Associate Executive Producer" credits, which can be officially listed on IMDb. A message from a donor in Yorkville, Illinois, alerts the community to a proposal for 14 new data centers in the small town, seeking advice on how to oppose the development. The segment reinforces the "Value for Value" model, thanking various supporters from Pittsburgh to India.