Coverage
6 Oct 2026
Super Intelligence (SI) news, papers, lab posts and video briefings from 6 Oct 2026. Every item links to its source.
Articles
AI doesn't need 'superintelligence' or evil intent to start a nuclear war
Capability · thebulletin.org
OpenAI safety leader quits, warning AI company's culture is 'broken'
Safety · theguardian.com
Forget 'superintelligence': error-prone AI nearly sparked world war three
Capability · theguardian.com
Administration unveils new Super Intelligence Force
Capability · techcrunch.com
Destroying all of humanity is hard work, even for a superintelligence
Capability · nibblestew.blogspot.com
Trump renames AI to 'super intelligence' and tells Americans they have no choice
Policy · independent.co.uk
Slovenia's .si domain sees surge in registrations after 'super intelligence'
Capability · techcrunch.com
AI Is Now Si: Super Intelligence Isn't Superior
Capability · tej.as
What I learnt co-leading an AI Safety bootcamp for legal and governance practit
Safety · lesswrong.com
Claude Ultracode is super intelligence
Labs · twitter.com
.si Registrations Surge After US Shift to 'Super Intelligence'
Capability · circleid.com
Trump's "super intelligence" naming problem
Policy · tech.yahoo.com
Please Wait, Super Intelligence Is Loading
Capability · magzimof.com
The AI boom is making the cheapest smartphones disappear
Capability · restofworld.org
Show HN: Moching - AI desktop agent with 219 built-in tools (Rust)
Capability · github.com
The AI Buildout Has Turned into a Doom Loop
Safety · bird.makeup
AI safety is mostly a sex cult in Berkeley, CA [pdf]
Safety · ams3.digitaloceanspaces.com
AI Backers Sound the Alarm About Safety
Safety · bloomberg.com
Show HN: AI Agents: Zero to Hero - Learn AI Agents from Scratch in Pure Python
Capability · github.com
Schema-guard, stop AI agents from inventing column names in SQL
Capability · github.com
Trump renaming AI to SI is making his wife's home country a little richer
Policy · neowin.net
SoftBank's Masayoshi Son has rare cautionary note on AI safety
Safety · japantimes.co.jp
The AI Safety community is unfortunately doing more harm than good
Safety · twitter.com
Japan's entertainment robots diversify with AI, offer comfort
Capability · mainichi.jp
Plainserp - Google search API for AI agents at $0.30 per 1k
Labs · plainserp.com
Wattle AI - an AI receptionist that answers and books calls for small businesses
Capability · heywattle.com
Can AI self-improvement overcome diminishing returns?
Capability · rameznaam.com
The ML4Good Technical AI Safety Bootcamp: What to Expect
Safety · lesswrong.com
How to apply to AI safety fellowships (and beyond)
Safety · lesswrong.com
Video briefings
Claude Opus 5.5 AI: An Incredible Leap Forward
Two Minute Papers · 359K views
From deepfakes to DNA: the science of watermarking AI
Google DeepMind · 63.1K views
Create your own voices with Gemini 3.8 text-to-speech
Google DeepMind · 139K views
Claude Is Now Leaving Invisible Fingerprints In Its Text
Two Minute Papers · 108K views
Gina Raimondo on the worst case scenario if 'we let AI loose'
CNBC Television · 2.8K views
Dazzle AI CEO: Our technology uses your photos to personalize suggestions
CNBC Television · 1.6K views
Papers
Second-Order Problem Solving for Recursive Self-Improvement in Formal Verification
arXiv:2610.05701 · Jiang, Y. et al.
Recursive Self-Improvement of Visuomotor Policies through Local Recovery Supervision
arXiv:2610.05151 · Zhang, Y. et al.
Assembling Insights for Agentic Machine Learning Engineering Systems
arXiv:2610.04927 · Jin, B. et al.
The Numerical Linear Algebra of Large Language Models
arXiv:2610.04631 · Baggag, A. et al.
What to Preserve in Recursive Computation: A Local Predictive Sufficiency Principle
arXiv:2610.04303 · Wang, P. et al.
From Valid to Useful: Post-Verification Acquisition for Recursive Self-Improving Recommendation
arXiv:2610.04302 · Hasan, T. et al.
EvalResearchBench: Can AI Agents Design Their Own Evaluations?
arXiv:2610.04184 · Zhang, Y. et al.
VIGOR: Zero-Shot Visual Generalization via Latent-Space Consistency in Model-Based Reinforcement Learning
arXiv:2610.02801 · Park, M. et al.
Decoding Looped Transformers Better for (Almost) Free
arXiv:2610.02185 · Liu, W. et al.
Can AI Oversight Be Zero Knowledge?
arXiv:2610.01995 · Chiesa, A. et al.
Safety Must Survive Self-Improvement: Why Failures Persist and How Agents Recover
arXiv:2610.01073 · Zhang, Y. et al.
Turbo Harness: Instance-Adaptive Harness Optimization
arXiv:2609.40330 · Zhang, T. et al.
Lab notes
Toward provably private learning from federated data
Google Research
The eternal complement
OpenAI