Podcast Summary

Podcast: Click Here (Recorded Future News)

Episode: Introducing Kill Switch

Release Date: December 16, 2025
Featured Show: Kill Switch (Kaleidoscope)
Host: Dexter Thomas
Main Guest: Dr. Corinne Kath, Culture Anthropologist (Article 19)

Overview

This episode of "Click Here" features an episode from the Kaleidoscope podcast "Kill Switch," hosted by Dexter Thomas. The focus is on a massive Amazon Web Services (AWS) outage that led to widespread disruption across the internet, impacting not just day-to-day conveniences but critical infrastructure like hospitals, banking, and government services. Through an in-depth conversation with Dr. Corinne Kath, the episode explains how the digital world is deeply reliant on a handful of large cloud providers, explores the history behind this consolidation, and considers the social, political, and safety implications of such concentrated power over foundational internet infrastructure.

Key Discussion Points & Insights

1. The AWS Outage: Scale and Impact

• Incident Recap:

  • In early November, a significant AWS outage knocked hundreds of services offline globally, including major websites, messaging apps, smart devices, and even government services. [01:11–02:33]
  • "This was huge. In my decade in the field, this definitely ranks in the top five."
    — Dexter Thomas [01:43]
  • Many users didn't realize a global outage was behind various disconnected services; they just noticed their favorite apps or smart gadgets stopped working.

• Widespread Consequences:

  • Examples included malfunctioning smart beds (e.g., "eight Sleep" mattresses), loss of access to apps like Signal, and failures of critical infrastructure.
  • "Thinking about somebody's $2,500 internet smart bed going vertical and trying to cook them alive is kind of dystopian, but it goes beyond that."
    — Corinne Kath [02:58]

• Critical Infrastructure at Risk:

  • Outages affect daily life, from banking to healthcare to government services.
    "This isn't just like me being on Duolingo or Snapchat, right? It is banking. It is the hospitals. It is education..."
    — Dexter Thomas [03:41]

2. The Anatomy of Cloud Dependency

• Only three major providers (Amazon, Microsoft, Google) account for almost two-thirds of the world's cloud infrastructure. If one suffers disruption, a significant portion of the Internet is affected. [03:58–04:38]

• Many apps and websites default to certain data centers (particularly the AWS Virginia cluster), further centralizing risk. [06:07]

  • "About 35 to 40% of all AWS traffic goes through it."
    — Dexter Thomas [06:07]

• Technical Explainer: The DNS Bug

  • The AWS outage was due to a Domain Name System (DNS) error in the Virginia data center.
  • Dexter uses a relatable metaphor: if your navigation app can’t find the library, you (and everyone else) can’t do your work—even though it’s still physically there. When the problem is fixed, there's a "stampede" as everyone reconnects, causing further backlog and delay. [07:13–09:45]

3. Repeated Cloud Outages: Patterns and Precedents

• Not the first major outage; previous large-scale outages affected Microsoft Azure (e.g., CrowdStrike update failure in 2024 that grounded airlines and impacted hospitals). [10:01–10:42]

  • "When you are in an ambulance being sped to a hospital, you don't want to show up there on the other end and them telling you I'm sorry, computer says no."
    — Corinne Kath [11:09]

• Such incidents are becoming disturbingly routine, yet infrastructure vulnerabilities remain unaddressed. [11:45–12:29]

• Even during their conversation, another cloud outage (Xbox, Minecraft, banks, train systems) was occurring, directly affecting elections in the Netherlands. [12:41]

  • "The outage was directly interfering with the democratic process."
    — Dexter Thomas [13:40]

4. How Did Amazon Become the Internet’s Backbone?

• Origin story: AWS began as a "side hustle," renting out Amazon's extra compute capacity.

• AWS solved a core risk for startups (scalability, cost flexibility), popularizing the "rent, not buy" model for servers and developer tools. [14:31–17:09]

  • "They were solving for a problem that companies were already starting to have... But it also meant that the Internet became really reliant on Amazon."
    — Corinne Kath [15:52]

• Interdependencies have grown complex; software is now a “web of building blocks,” with each block relying on others, deepening systemic risk. [17:09]

• AWS now accounts for more than half of Amazon's operating income. Google and Microsoft followed, leading to the current oligopoly. [17:47]

• Notable Metaphor:
  "In what situation would you accept only three companies being responsible for all of agriculture?... We need to start changing our perception of what the digital is because it is everything."
  — Dexter Thomas [18:20]

5. Monopoly vs. Resilience: Economic and Social Arguments

• The cost and frequent updates required for cloud infrastructure create high barriers to competition. VCs are unwilling to fund alternatives to the Big Three. [19:29–20:39]

  • "It's a bad investment right now."

• Some argue "big is better" for reliability (resources, scale), but Corinne sees political and societal dangers in this monopolization. [20:39–22:03]

  • "It's very easy for people to say, 'Oh, it's just a technical mistake and we should be allowed to make technical mistakes.' But that is not an argument we would accept from the company that operates our water or electricity..."
    — Dexter Thomas [21:00]

• Outages are not just technical failures—they have major political and individual consequences and illustrate a “democratic deficit.”

6. Rethinking Digital Sovereignty and Possible Alternatives

• European leaders are debating "digital sovereignty," seeking to reduce dependence on U.S. cloud giants. But non-monopolistic, globally fair alternatives are needed (not just defense contractors). [24:48–26:06]

• Non-profits and universities have successfully built distributed, independent systems (e.g., Wikipedia and Dutch universities collaborating on university-specific cloud infrastructure). [26:14–27:23]

• Historically, the ARPANET (precursor to the Internet) was intentionally designed without single points of failure, unlike today’s monopoly-driven structure. [27:31–28:12]

• Promising Projects:

  • University of Amsterdam's Critical Infrastructure Lab and Brussels’ Institute for Technology in the Public Interest: exploring radical alternatives like compostable data centers and “cloud abolitionist” agendas. [28:42–29:51]
  • "We need people who will actually question the premises and not just try and improve what we have already."
    — Dexter Thomas [29:51]

7. The Future: Is This Problem Getting Worse?

• Outages are happening frequently and affecting more essential sectors as more critical infrastructure (e.g., trains, healthcare) moves to cloud. [29:51–30:53]

  • "Are we likely to see this more in the future? Absolutely. Will the impact be bigger every time? Yes..."
    — Dexter Thomas [30:08]

• Example: Even everyday conveniences (like using a phone to pay for LA transit) have high-stakes consequences if the cloud fails (risk of police harassment, etc.). [30:53]

8. Who Should Decide What We Can Access Online?

• Highlight of AWS’s decision to deplatform Parler after Jan. 6th as an example of private power controlling public discourse. [31:24–32:30]

  • "What this has shown us is the specific power dynamic of how these cloud companies get to decide who gets to use their services and by extension, who gets to exercise their rights online."
    — Dexter Thomas [32:30]

• Critical communication tools (like Signal) rely on AWS. What if AWS withdrew support due to policy or politics? [33:17–33:40]

• The Accountability Question:

  • Holding cloud giants accountable is challenging for individuals, but legal and political advocacy (such as lawsuits over unfair licensing) is possible. [33:40–34:40]
  • "The entirety of the weight of that corporate power comes bearing down on you. So you need to have the kind of backbone to take on that backbone."
    — Dexter Thomas [34:40]

• At minimum, ask institutions about contingency plans: "Not if, but when that [cloud] service goes offline next."
  — Corinne Kath [34:40–35:32]

Notable Quotes & Memorable Moments (with Timestamps)

• "This was huge. In my decade in the field, this definitely ranks in the top five."
  — Dexter Thomas [01:43]
• "Thinking about somebody's $2,500 internet smart bed going vertical and trying to cook them alive is kind of dystopian, but it goes beyond that."
  — Corinne Kath [02:58]
• "When you are in an ambulance being sped to a hospital, you don't want to show up there on the other end and them telling you I'm sorry, computer says no."
  — Corinne Kath [11:09]
• "In what situation would you accept only three companies being responsible for all of agriculture?... We need to start changing our perception of what the digital is because it is everything."
  — Dexter Thomas [18:20]
• "It's very easy for people to say, 'Oh, it's just a technical mistake and we should be allowed to make technical mistakes.' But that is not an argument we would accept from the company that operates our water or electricity..."
  — Dexter Thomas [21:00]
• "What this has shown us is the specific power dynamic of how these cloud companies get to decide who gets to use their services and by extension, who gets to exercise their rights online."
  — Dexter Thomas [32:30]
• "The entirety of the weight of that corporate power comes bearing down on you. So you need to have the kind of backbone to take on that backbone."
  — Dexter Thomas [34:40]

Key Segment Timestamps

| Segment | Timestamp | |---------------------------------------------------------------------|------------------| | Outage recap and AWS’s criticality | 01:11–03:41 | | Dependency on AWS and the AWS Virginia data center | 05:07–07:13 | | DNS error metaphor and impact explanation | 07:13–09:45 | | Past outages: Microsoft Azure, real-world hospital risks | 10:01–12:29 | | Cloud monopoly origins and market structure | 14:31–17:47 | | Monopoly critique: infrastructure as public good | 17:47–22:03 | | European digital sovereignty, alternatives to cloud monopoly | 24:48–27:23 | | ARPANET history and lessons for modern Internet design | 27:31–28:28 | | Cloud accountability, AWS’s content moderation power | 31:24–32:30 | | Calls to action, systemic vs. individual solutions | 33:53–35:32 |

Tone & Style Notes

• Dexter Thomas uses everyday analogies, humor, and accessible metaphors to explain technical concepts.
• Dr. Corinne Kath blends technical clarity with frank social and political commentary.
• The episode maintains an urgent, conversational, and slightly irreverent but deeply informed tone throughout.

Takeaway

The episode brings into sharp relief how invisible yet essential internet infrastructure has become to every aspect of modern life—and the risks society faces by concentrating so much control in so few hands. Outages are not rare, and as dependency deepens, the impact grows. The solution is not solely technical: it demands rethinking digital power, pushing for political accountability, and imagining alternative models rooted in public interest and resilience.