The 404 Media Podcast: "The Depravity Economy"

Episode Date: March 4, 2026
Hosts: Jason Kebler, Emmanuel Maiberg, Samantha Cole, Matt Gault
Main Themes: The rise of prediction markets in geopolitics and finance ("the depravity economy"); how AI-generated translations are polluting Wikipedia; the weaponization of data centers in modern conflict.

Episode Overview

This episode of 404 Media dives into two headline investigations: the controversial rise of prediction markets like Kalshi and Polymarket in the context of recent geopolitical strife, and the infiltration of AI-powered translation errors into Wikipedia articles. The hosts also touch on the recent Iranian missile strikes that damaged Amazon data centers and what this means for future warfare and civilian harm.

Segment 1: Prediction Markets and the "Depravity Economy"

[01:24–24:00]

Introduction to Prediction Markets

• Definition & Context: Apps like Kalshi and Polymarket allow anyone to bet on real-world outcomes—elections, commodity prices, or even the outcomes of wars.
• Origin in sports betting: Daily fantasy services like DraftKings/FanDuel normalized rapid, granular betting.
• Expansion: These platforms have extended to permit bets on anything, including highly sensitive geopolitical events.

"They are these apps and websites that allow you to basically bet on the outcome of anything that happens, like in the world, essentially."
—Jason, [02:19]

Controversies and Ethical Questions

• Manipulation Risks:
  Matt recounts a Polymarket bet about the Ukraine-Russia war that was manipulated by someone editing an ISW conflict map to influence the bet outcome.

  • "It appears that someone at ISW who had access to edit the map, laid a bet and altered the map... made it appear that Russia had taken a town square and collected their money."
    —Matt, [06:54]

• Arbiter Challenges: Bets require a trusted third party to declare results, exposing weaknesses in verifiability.

• Profiting from Death:
  Recent Iran bets involved speculation on whether the Ayatollah would remain in power. Despite his death in an airstrike, Kalshi refused to honor payouts, citing a prohibition on directly profiting from death.

  • "Kalshi's CEO... said, like, we are not paying this bet out because we don't allow people to directly profit from death."
    —Jason, [09:38]
  • Kalshi refunded bets, sparking community backlash.

• Arbitrary Distinctions:
  Despite banning assassination bets, users can still gamble on metrics tightly linked to violence and civilian harm (e.g., gas prices, military strikes).

  • "Any way you slice it, it's betting on death."
    —Jason, [14:38]

The "Depravity Economy" Concept

• Definition: The normalization of profiting from all forms of vice and tragedy, driven by both the collapse of traditional pathways to prosperity and the relentless financialization of everyday life.
  • Examples: Meme coins, AI gold rush, scammy financial hustles, even meme stocks like Palantir, where profit hinges on human rights abuses or war.
  • "All vices are game at this point... like human trafficking, like, who knows?"
    —Jason, [18:40]
• Rapid Mainstreaming: Advertisements and sponsorships rapidly legitimize and entrench these gambling markets across media and sports.
  • "Polymarket and Kalshi are like on the broadcast [at the Super Bowl]. If you go to a sporting event, like, their names are on the court and things like this."
    —Jason, [18:39]

Prediction Markets as "News"

• Polymarket's CEO markets his platform as the future of news, arguing prediction market probabilities better capture reality than traditional reporting.
  • "He thinks that Polymarket is a better way to deliver news to people, it is the future of news."
    —Matt, [20:55]

Notable Quote

"Perverse incentives create perversion, right?"
—Matt, [08:45]

Segment 2: Data Centers as Targets in Modern Conflict

[24:00–28:36]

Iranian Strikes Hit Amazon Data Centers

• Impact: Amazon Web Services data centers in Dubai and Bahrain knocked offline after missile/drone strikes.
  • Initial clue: social media complaints about service outages, official AWS health dashboard updates, and confirmation from Amazon.
  • "Yesterday at 7 eastern time... it was Iranian drone strikes on, on two data centers in Dubai."
    —Matt, [24:23]
• Broader Implications: Targets no longer limited to traditional military assets—striking civilian/commercial cloud infrastructure can cause widespread disruption.
  • Matt: "If I can knock out the system that's going to help you do targeting or help you run missile defense, I'm gonna do it, it'll be the first thing I attack, you know. So, yeah, absolutely. I think data centers are targets, for sure." [27:08]
  • Parallels drawn to Russian attacks on the Ukrainian energy grid.

Segment 3: AI Translation Hallucinations Pollute Wikipedia

[33:18–53:47]

Wikipedia's Nuanced AI Policy

• Wikipedia's volunteers are highly wary of AI-generated errors but don't impose blanket bans. Instead, they impose strict review and block repeat offenders.
  • "Zero tolerance for AI generated errors."
    —Emmanuel, [34:20]
• Task forces actively seek out and fix AI mistakes.

The Open Knowledge Association (OKA) Scandal

• OKA's Project: Pays translators (often in the global South, for low compensation) to use AI to generate translated Wikipedia drafts.

  • High quotas may incentivize errors.

• Major Problems Identified:

  • Hallucinated Citations: AI fabricates references or links that don’t exist.
  • Factual and Formatting Errors: Copy-pasting from LLMs leads to subtle or overt mistakes.
  • Use of Grok, an LLM that once made a Wikipedia competitor (Grokopedia), further irks volunteers due to prior data scraping and poor factual quality.

• AI Checking AI:
  Founder Jonathan Zimmerman proposes using one LLM to verify another, a "flawed method," hosts note, since it's prone to missing the same mistakes.

Editorial Response:

Wikipedia editors target specific offenders instead of banning OKA or AI tools outright. Repeat AI-related errors lead to stricter account reviews or bans. - "There's like a two strikes and you're out kind of rule. So if an editor uses AI translation twice and they're given a warning, but they find two drafts that have AI generated errors, they prevent the account from making contributions."
—Emmanuel, [44:25]

The Broader Debate Over AI in Translation

• Discusses HarperCollins using AI to translate Harlequin novels—cheaper and faster, but only with substantial human correction.

• AI is useful for gist and information gathering (e.g., news research), but always needs a human check for high-stakes content.

  • "LLMs... are extremely good at translating from one language to another, can be a very useful tool. Crucially in both cases I was able to check it with somebody who actually speaks the language, which is what you should be doing."
    —Emmanuel, [50:18]

• On artistic translation (romance novels, anime, games), the group worries AI will only serve "slop" to low-information readers while quality-seekers may migrate elsewhere.

Segment 4: Closing Thoughts & Notable Quotes

• The hosts reflect on the general degradation and "sloppification" of formerly serious mediums—finance, news, art—by financialization and AI.
• The market incentives driving sloppier, faster, and more ethically dubious practices are unlikely to abate soon.
  • "The market for those books is like, can you trick people into buying them on accident?"
    —Jason, [49:27]

Key Quotes and Memorable Moments

• "This is reaching sort of the culmination where... you can make tons of money if you are super up on the news or have some sort of inside information and you bet on just like death and destruction and human suffering."
  —Jason, [18:40]
• "Nothing of value is being created here."
  —Jason, [23:56]
• "All vices are game at this point."
  —Jason, [18:40]
• "Perverse incentives create perversion."
  —Matt, [08:45]
• "Crucially in both cases I was able to check it with somebody who actually speaks the language, which is what you should be doing."
  —Emmanuel, [50:18]

Notable Timestamps

• [01:24] Introduction to prediction markets
• [06:54] Example of market manipulation and perverse incentives
• [09:38] Iran/Ayatollah market controversy
• [17:15] Explanation of "depravity economy"
• [24:23] Amazon data centers and impact of strikes
• [34:20] Wikipedia’s nuanced stance on AI
• [41:31] LLM-generated content infecting Wikipedia
• [44:25] Wikipedia's rule for AI abuse prevention
• [50:18] Practical AI translation experiences

Overall Tone:
The conversation is skeptical, at times sardonic, but grounded in deep expertise—especially in untangling the real and philosophical consequences of rapid tech disruption. The hosts are witty, self-aware, and keenly concerned about how financial incentives and AI are reshaping society’s relationship to truth, art, and even mortality.