RoundTalk

A viral AI 'doomsday' warning went viral: should we really be scared?

English · B1 worksheet Based on the RoundTalk episode of

Free for classroom use. Students listen to the first 75 seconds and read the beginning of the debate, then work on the tasks. The answer key is on the last page. The panelists are fictional characters with AI voices.

Episode page
1 · Listen (first 75 seconds)

2 · Read

Host

Hello and welcome. So, a former AI researcher quit his job and told the world that the labs building superintelligence might kill us all—and here's the thing, the CEOs didn't really push back. That's sparked the biggest doomsday conversation in tech. But is this genuine alarm or a clever move by AI companies to shape regulation? That's what we're asking today: should we take Jacob Coxon's warning seriously, or is it a self-serving narrative pushed by labs to justify the rules they want? To dig into this, I've got Marcus Chen with us, an AI safety researcher who's spent twelve years studying alignment and control. He knows the technical side cold. Sophie Okonkwo joins us too—she's a tech policy journalist who's been tracking corporate lobbying and incentives in Silicon Valley for eight years, so she follows the money behind these stories. And David Rothstein, who advises startups and nonprofits on safety practices and used to lead ethics at a tech company, he brings the practical angle—what can teams actually do right now.

Host

Marcus, let's start with you. When Coxon talks about extinction risk, what's actually plausible here?

Marcus Chen

Let me be precise here. Extinction is not the plausible part. Losing control of fast, self-improving systems is. Hubinger says over ten percent in ten years. That number is a guess, honestly.

Sophie Okonkwo

Sorry, but that guess helps them get the rules they want.

Marcus Chen

No, it's the other way round. That ten percent number hurts them. It invites rules they hate. The thing is, I was writing about control failures in 2018, when nobody cared.

Host

David, you work with teams building AI right now. When researchers talk about losing control, what does that actually mean in a lab?

Key vocabulary

  • superintelligence — artificial intelligence more powerful than humans
  • resignation — act of quitting a job voluntarily
  • extinction — complete destruction or elimination
  • surveillance — close watch or monitoring of people
  • doomsday — a day of final disaster or ruin

Comprehension questions

  1. What does Marcus Chen say is the plausible danger, not extinction?
  2. Why does Sophie Okonkwo think the ten percent number helps AI companies?
  3. What does Marcus say the ten percent number actually does to AI companies?
  4. When did Marcus start writing about control failures?
  5. What is the main question the debate wants to answer?

True or false?

  1. Marcus Chen has spent eight years studying AI alignment and control. T / F
  2. Sophie Okonkwo has been tracking corporate lobbying and incentives in Silicon Valley for eight years. T / F
  3. Marcus thinks the ten percent number helps AI companies get the rules they want. T / F
  4. Jacob Coxon is an AI researcher who quit his job and warned about superintelligence. T / F

Fill the gaps

doomsdayextinctionsuperintelligence
  1. So, a former AI researcher quit his job and told the world that the labs building   might kill us all—and here's the thing, the CEOs didn't really push back.
  2. That's sparked the biggest   conversation in tech.
  3. When Coxon talks about   risk, what's actually plausible here?

Discuss in pairs

  1. Do you think AI companies are really scared of losing control, or are they using safety fears to get the government to make the rules they want?
  2. If AI researchers were writing about control problems in 2018 but nobody listened, why do you think people care more now?
  3. What should teams building AI do right now to be more careful, in your opinion?

The whole debate, free every day Students can hear the full 6-minute episode with a synced transcript in the RoundTalk app.

Get it on Google Play Download on the App Store

Answer key

Comprehension
  1. Losing control of fast, self-improving systems.
  2. Because it helps them get the rules they want.
  3. It hurts them because it invites rules they hate.
  4. In 2018, when nobody cared.
  5. Whether we should take Jacob Coxon's warning seriously or if it's just about regulation.
True or false
  1. False
  2. True
  3. False
  4. True
Gaps
  1. superintelligence
  2. doomsday
  3. extinction

roundtalk · https://www.getroundtalk.com/listen/english/b1/a-viral-ai-doomsday-warning-went-viral-should-we-really-be-scared/worksheet/