<!-- mobian-agent-page publisher="dailydev" canonical="https://daily.dev/posts/DUCtcwiyd" -->

---
title: They really do think AI might kill everyone | daily.dev
description: A viral resignation tweet from an Anthropic researcher claiming AI could kill everyone by the end of the decade reignited debate about AI doom scenarios....
canonical: https://daily.dev/posts/they-really-do-think-ai-might-kill-everyone-ductcwiyd
twitter:card: summary_large_image
twitter:site: @dailydotdev
og:type: website
og:site_name: daily.dev
og:title: They really do think AI might kill everyone | daily.dev
og:description: A viral resignation tweet from an Anthropic researcher claiming AI could kill everyone by the end of the decade reignited debate about AI doom scenarios....
og:url: https://daily.dev/posts/they-really-do-think-ai-might-kill-everyone-ductcwiyd
og:image: https://api.daily.dev/og/posts/DUCtcwiyd.png
og:image:alt: They really do think AI might kill everyone
og:image:width: 1200
og:image:height: 630
og:locale: en
---

> ## Documentation Index
> Fetch the complete documentation index at: https://daily.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# They really do think AI might kill everyone

**[Tech Lead Digest](https://daily.dev/sources/techleaddigest)** · 10 min read · 1 upvotes · 0 comments

## Summary

A viral resignation tweet from an Anthropic researcher claiming AI could kill everyone by the end of the decade reignited debate about AI doom scenarios. Rather than dismiss such views as PR, self-promotion, or stock manipulation, this piece traces the AI safety community's long history (since the mid-2000s, with figures like Eliezer Yudkowsky) of sincerely believing superintelligent AI poses existential risk. It walks through mechanisms researchers propose - engineered pandemics, AI-triggered nuclear war, robotics takeover, nanotech 'grey goo' - and rebuts common counterarguments (partial extinction still counts, governments can't simply nationalize labs, 'just turn it off' won't work against a sufficiently capable AI). It also explains why believers keep working in AI rather than opposing it: many think only aligned superintelligence can save humanity long-term, and 'foom'-style takeoff dynamics mean whoever gets there first can halt rival efforts.

## Full article

daily.dev links to this article rather than hosting it. Read it at the original source: <https://www.seangoedecke.com/they-really-do-think-ai-might-kill-everyone>

## Questions this post answers

### What are the main ways researchers think a superintelligent AI could cause human extinction?

Proposed mechanisms include an AI engineering and releasing a maximally lethal pathogen via autonomous biology labs, triggering global thermonuclear war by infiltrating military or government decision systems, seizing robotics and drones to physically eliminate people, or using nanotechnology to create self-replicating 'grey goo.' These scenarios are ranked roughly in order of plausibility by AI safety researchers, with bioweapons considered most concerning given current autonomous lab development.

_Developers tracking AI safety debates can follow this reasoning and its critiques on daily.dev._

### Why do AI safety researchers keep building AI if they believe it could cause extinction?

Many believe an AI powerful enough to end humanity is also powerful enough to save it, so achieving aligned superintelligence is seen as necessary for humanity's long-term survival. There's also a race dynamic: under 'foom' or hard-takeoff theories, the first lab to crack self-improving AI could stop all rival AI research by hacking, persuading, or disabling competing labs, making it seem safer to get there first than to abstain.

_Anyone weighing the ethics of AI development work can dig into this reasoning on daily.dev._

### What is p(doom) in AI safety discussions?

P(doom) is shorthand, in use since around 2010 within the AI research community, for the probability that AI causes human extinction or an equivalently catastrophic outcome. The term traces back to work by figures like Eliezer Yudkowsky, who has published on superintelligent AI risk since at least 2008, well before mainstream AI discourse treated the topic seriously.

_Readers gauging how seriously to take AI risk framing can explore the debate on daily.dev._

## Similar posts on daily.dev

- [The AI doomers feel undeterred](https://daily.dev/posts/the-ai-doomers-feel-undeterred-5elrg9q7q) · MIT Technology Review · 1 upvotes · 0 comments
- [Yoshua Bengio warns hyperintelligent AI with](https://daily.dev/posts/yoshua-bengio-warns-hyperintelligent-ai-with-2qrxdw7fa) · The Next Web · 0 upvotes · 0 comments
- [Will This Creature Destroy Its Creator?](https://daily.dev/posts/will-this-creature-destroy-its-creator--nkjykoftt) · Medium · 0 upvotes · 0 comments

---

Tags: [#ai](https://daily.dev/tags/ai), [#anthropic](https://daily.dev/tags/anthropic), [#ai-safety](https://daily.dev/tags/ai-safety)

[View this post on daily.dev](https://daily.dev/posts/they-really-do-think-ai-might-kill-everyone-ductcwiyd)

```json
{"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]}
{"@context":"https://schema.org","@type":"TechArticle","headline":"They really do think AI might kill everyone","url":"https://daily.dev/posts/they-really-do-think-ai-might-kill-everyone-ductcwiyd","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/they-really-do-think-ai-might-kill-everyone-ductcwiyd"},"datePublished":"2026-09-13T21:43:36.366Z","dateModified":"2026-09-13T21:44:01.075Z","description":"A viral resignation tweet from an Anthropic researcher claiming AI could kill everyone by the end of the decade reignited debate about AI doom scenarios....","image":"https://media.daily.dev/image/upload/s--P4t4XyoV--/f_auto/v1722860399/public/Placeholder%2001","thumbnailUrl":"https://media.daily.dev/image/upload/s--P4t4XyoV--/f_auto/v1722860399/public/Placeholder%2001","isAccessibleForFree":true,"articleSection":"Tech Lead Digest","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Tech Lead Digest","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/6ed1ee1a2e5c49f698265a6b2912b80d","url":"https://daily.dev/sources/techleaddigest"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/they-really-do-think-ai-might-kill-everyone-ductcwiyd","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai,anthropic,ai-safety","timeRequired":"PT10M"}
{"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Tech Lead Digest","item":"https://daily.dev/sources/techleaddigest"},{"@type":"ListItem","position":3,"name":"They really do think AI might kill everyone"}]}
{"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/they-really-do-think-ai-might-kill-everyone-ductcwiyd#faq","mainEntity":[{"@type":"Question","name":"What are the main ways researchers think a superintelligent AI could cause human extinction?","acceptedAnswer":{"@type":"Answer","text":"Proposed mechanisms include an AI engineering and releasing a maximally lethal pathogen via autonomous biology labs, triggering global thermonuclear war by infiltrating military or government decision systems, seizing robotics and drones to physically eliminate people, or using nanotechnology to create self-replicating 'grey goo.' These scenarios are ranked roughly in order of plausibility by AI safety researchers, with bioweapons considered most concerning given current autonomous lab development. Developers tracking AI safety debates can follow this reasoning and its critiques on daily.dev."}},{"@type":"Question","name":"Why do AI safety researchers keep building AI if they believe it could cause extinction?","acceptedAnswer":{"@type":"Answer","text":"Many believe an AI powerful enough to end humanity is also powerful enough to save it, so achieving aligned superintelligence is seen as necessary for humanity's long-term survival. There's also a race dynamic: under 'foom' or hard-takeoff theories, the first lab to crack self-improving AI could stop all rival AI research by hacking, persuading, or disabling competing labs, making it seem safer to get there first than to abstain. Anyone weighing the ethics of AI development work can dig into this reasoning on daily.dev."}},{"@type":"Question","name":"What is p(doom) in AI safety discussions?","acceptedAnswer":{"@type":"Answer","text":"P(doom) is shorthand, in use since around 2010 within the AI research community, for the probability that AI causes human extinction or an equivalently catastrophic outcome. The term traces back to work by figures like Eliezer Yudkowsky, who has published on superintelligent AI risk since at least 2008, well before mainstream AI discourse treated the topic seriously. Readers gauging how seriously to take AI risk framing can explore the debate on daily.dev."}}]}
```

