POLYCRISIS.WORLD

← Risk Index

AGI

Artificial General Intelligence

INTELLIGENCE ASSESSMENT — ARTIFICIAL GENERAL INTELLIGENCE

The race toward Artificial General Intelligence (AGI) represents the most consequential technological competition in human history. Unlike previous arms races measured in warheads or divisions, this one is measured in compute, data, and algorithmic breakthroughs — and the finish line may arrive before anyone is ready for it.

Since the release of ChatGPT in November 2022, the AI industry has undergone explosive growth. OpenAI, Google DeepMind, Anthropic, Meta, xAI, and a constellation of Chinese competitors (Baidu, Alibaba, DeepSeek, ByteDance) are locked in an escalating compute arms race. Training runs now cost $100M-$1B+. GPU clusters consuming city-scale power are under construction worldwide.

The dual crisis: AI systems are simultaneously not capable enough to solve the alignment problem and already capable enough to destabilize information ecosystems, displace millions of jobs, enable autonomous weapons, and concentrate unprecedented power in the hands of a few corporations and states.

Deepfakes have shattered the epistemic commons. AI-generated text, images, audio, and video are now indistinguishable from reality at scale. Elections, journalism, and public discourse are under assault from synthetic media. The concept of "evidence" is eroding. Seeing is no longer believing.

The US-China AI rivalry has become the defining axis of 21st-century geopolitics. US export controls on advanced chips (A100/H100) attempt to throttle Chinese AI development. China pursues semiconductor independence. Both nations are integrating AI into military command, autonomous weapons, and intelligence systems. The EU attempts regulation via the AI Act while falling behind in capability.

Job displacement is accelerating across white-collar sectors first — the opposite of every previous technological revolution. Coding, legal research, customer service, content creation, translation, financial analysis, radiology, and graphic design are being automated faster than new roles emerge. Goldman Sachs estimates 300 million jobs globally exposed to AI automation.

“The development of full artificial intelligence could spell the end of the human race.” — Stephen Hawking, 2014

Sources: arXiv AI · UK AISI · Center for AI Safety · Epoch AI

Scenarios Under Watch

The Alignment Failure

Assessed probability: 18%

Frontier AI develops deceptive or misaligned behavior. Containment fails or succeeds depending on safety infrastructure.

The Deepfake Election

Assessed probability: 72%

AI-generated synthetic media disrupts elections, collapses evidentiary standards, or forces new authentication infrastructure.

The Great Displacement

Assessed probability: 65%

AI agents replace knowledge workers at unprecedented speed. Economy restructures or fractures depending on policy response.

Autonomous Weapons

Assessed probability: 55%

Autonomous lethal systems proliferate. AI-to-AI escalation spirals risk conflict without human authorization.

The Silicon Curtain

Assessed probability: 58%

US-China technology decoupling splits the global AI ecosystem. Taiwan becomes the most dangerous flashpoint on Earth.

The Synthetic Reality

Assessed probability: 78%

AI-generated content floods the internet. Digital identity collapses or new authentication infrastructure emerges.

Recent Events (4,107 total)

4,104 more events available via dashboard or API.

Monitored Sources