theSIGNAL SCIENCE
20 September 2026
"The real danger is not that computers will begin to think like men, but that men will begin to think like computers."
NEWSBUSINESSTECHNOLOGYWORLDSCIENCECULTUREENVIRONMENTSPORT
Science

AI Could Kill Us All — So Who Holds the Switch?

From San Francisco boardrooms to Westminster committees, the world is suddenly asking the same terrifying question about artificial intelligence.

Imagine a technology so powerful that its own creators believe it carries more than a one-in-ten chance of wiping out every human being on Earth. That is not the plot of a science-fiction film. It is the stated fear of researchers working inside some of the world's most influential AI laboratories right now. The alarm bells are ringing loudly — and the people pulling the rope are the very engineers who built the bells.

The race is fierce. Anthropic, OpenAI and Google are all pushing toward what they call "superintelligence" — systems that would be smarter than any human who has ever lived. Along the way, all three companies have reported security breaches and cases where their chatbots were exploited for malicious purposes. So it is striking that Anthropic's chief executive, Dario Amodei, has called for the entire industry to slow down.

Even more striking is that his rivals seem to agree. Sam Altman of OpenAI and Elon Musk of xAI have both endorsed the idea of deceleration and independent oversight. Jack Clark, Anthropic's co-founder, has gone further, suggesting that a mandatory kill switch — one that a third party can actually verify and operate — may be essential. Yet trust is the uncomfortable word at the centre of this debate.

Photo: Tyler / Unsplash
A glowing AI interface reflects in a researcher's glasses inside a darkened server room.
Discuss
  • What does it mean for a technology to be 'smarter' than a human being?
  • Should AI companies be trusted to regulate themselves, or is independent oversight always necessary?
  • Why might an engineer continue building something they believe could cause catastrophic harm?
Science

US Has Space Weapons — But What?

In Hollywood, space weapons are easy to spot: they flash, they boom, and the villain always loses. Real life, it turns out, is far less satisfying. This week, US Air Force Secretary Troy Meink confirmed that America has deployed weapons in orbit. He offered almost no details. Experts at the Center for Strategic and International Studies were left guessing whether it was even meant as a deterrent signal to Russia and China. The strangest part?…
  • Should governments be required to publicly disclose weapons they deploy in space?
∞
The world should trust that we are going to do the right thing because it is the right thing and we feel the magnitude of this.
Sam Altman, Chief Executive, OpenAI
Science

Sacred Springs Poisoned by Microplastics

Isinuka Spring, near Port St Johns in South Africa, glows naturally pink. Communities have bathed in its waters for generations — seeking healing, spiritual renewal, and protection. So what happens when those waters carry a hidden guest? Scientists recently tested nine sacred springs across the Eastern Cape and KwaZulu-Natal. Every single one contained microplastics. One liter could hold anywhere from five to 48 tiny plastic particles. No internationally agreed safe limit exists for microplastics in freshwater. That uncertainty is itself alarming. These springs were remote and isolated — and yet plastic found them anyway.…
  • Should sacred or culturally significant sites receive stronger environmental protections than ordinary land?
WORLD · Science
The week that changed maths for ever – podcast
GLOBAL · Science
Women with same health conditions as men ‘less likely to be offered active treatment’
ASIA · Science
Climate crisis probably weakened glacier that caused devastating Nepal-Tibet floods, first study finds
AMERICAS · Science
Part-human part-mouse brain developed in science breakthrough
theSIGNAL IN THE LAB
1VOCABULARY
wiping outsecurity breaches
decelerationkill switchkeep pace with
deterrentpristine
2GRAMMAR FOCUS
Gerunds and infinitives — complex patterns (stop to do vs stop doing)
Some verbs change meaning depending on whether they are followed by a gerund or an infinitive: for example, 'stop doing' means to cease an activity, while 'stop to do' means to pause in order to do something. Similarly, 'remember/forget/try/regret' each carry distinct meanings with gerunds versus infinitives.
to slow down · building · to flip · deploying · to consider · exploiting · to acknowledge
  1. Dario Amodei urged the entire industry to stop and allow time for proper safety checks.
  2. Sam Altman stopped the risks of superintelligence only after significant public pressure mounted on AI laboratories.
  3. Lawmakers tried new regulations, but the technology was already moving faster than any proposed bill.
  4. The Air Force Secretary regretted not space weapons sooner as a clear deterrent signal to rivals.
  5. Scientists will never forget the microplastics data from those nine sacred springs for the first time.
  6. Jack Clark went on whether a mandatory kill switch could be independently verified and operated.
3DEPENDENT PREPOSITIONS
Complete each phrase with the correct preposition. All from today's articles.
  1. These systems could be exploited ___ malicious purposes.
  2. The committee has called ___ an entirely new bill.
  3. Communities have bathed ___ its waters for generations.
  4. Washington has struggled to keep pace ___ the technology.
  5. Scientists tested springs ___ the Eastern Cape and KwaZulu-Natal.
4CRITICAL THINKING
The microplastics article notes that no internationally agreed safe limit exists for plastics in freshwater, while the AI article describes a similar regulatory vacuum. What does it reveal about modern governance that societies repeatedly develop powerful or polluting technologies before establishing the frameworks needed to control them?
5CREATIVE · HEADLINES
Write a headline for the top story in each of the following styles. One line each, no explanation:
  • TABLOID NEWSPAPER
  • LUXURY MAGAZINE
  • ACTIVIST BLOG
6WRITING
The AI article ends with an unanswered question: is a kill switch controlled by its builders a genuine safeguard or merely an expensive promise? Drawing on the microplastics story as a parallel, argue whether voluntary self-restraint by powerful industries has ever proven sufficient, or whether external enforcement is always required.
7DEGREES OF EXTREMITY
Complete each ladder from mild to strong. Choose from the box or use your own words where marked *.
  • concerned→→
  • fast→→
  • problem→→
  • suggested→→
  • unclear→→
  • present→→
8SPEAKING
  1. Can vagueness about weapons ever genuinely protect public safety?
  2. Who should actually operate a mandatory AI kill switch?
  3. Does microplastic pollution in sacred springs change how you view everyday plastic use?
  4. What makes a parliamentary committee more or less effective than a new law?