Top news
Investing in the public benefit OfficialAnthropic NewsOct 18Popular Musk Pledges SpaceX Shares to ‘Every Kid in America’ WebBloomberg TechnologyOct 19Popular Websites are closing their doors to AI agents WebThe Verge AIOct 17 This is the new ChatGPT Voice, powered by GPT Live OfficialOpenAI on YouTubeOct 18 How AI giants will reshape global philanthropy WebAxios AIOct 19 Gemini 4 Argon Really Surprised Me WebTwo Minute PapersOct 19 Google Gemini 4 agents just beat Minecraft with no human intervention WebEngadget AIOct 19 Who could it be? 🧐 OfficialGoogle on YouTubeOct 18 Anthropic gives outside AI watchdog “employee-like” access WebArs Technica AIOct 18 What are self-replicating ‘AI worms’, and why is Anthropic worried about them? WebThe Guardian AIOct 18
GlossaryModelsIntermediate

Artificial general intelligence

Also written: AGI, Physical AGI, Embodied AGI, AGI future, Artificial General Engineer, AGI moment, AGI milestone, AGI Preparedness

AI that can think and learn like a human across many different jobs.

In plain terms

Artificial general intelligence, or AGI, is a type of AI that could perform any intellectual task a human can. Think of it as an AI that can understand, learn, and apply knowledge broadly, not just in one specific area.

Right now, most AI systems are narrow, meaning they are trained for a specific task, like playing chess or translating languages. AGI would be different. It would be able to switch between tasks, reason about new problems, and learn new skills without being explicitly reprogrammed for each one. For example, an AGI could potentially learn to cook, write a novel, and then help design a bridge, all using its general intelligence.

Achieving AGI is a major goal for many AI labs. It represents a significant leap beyond current AI capabilities and raises important questions about the future of technology and society.

The expert version

Artificial general intelligence (AGI) refers to a hypothetical AI system possessing human-level cognitive abilities across a wide spectrum of tasks. Unlike narrow AI, which excels at specific functions (e.g., image recognition, natural language processing), AGI would demonstrate adaptability, reasoning, and learning capabilities applicable to novel and diverse challenges. The development of AGI is often framed as achieving consciousness or sentience, though a precise, universally agreed-upon definition and measurable criteria remain subjects of ongoing debate and research.

Current research pathways towards AGI explore various architectures and methodologies, including large language models, reinforcement learning, and neuro-symbolic approaches. The core challenge lies in replicating the flexibility, common sense, and abstract reasoning that characterize human intelligence. Tradeoffs involve balancing performance on specific tasks with the generality and adaptability required for true AGI.

Why it matters now

Recent discussions and potential demonstrations, like with Anthropic's Claude Opus 5.5, are reigniting the conversation around Artificial General Intelligence. Companies like OpenAI and Google DeepMind are actively engaging in building standards and widening the debate, indicating a renewed focus on the path towards human-level AI capabilities.

Where it shows up 132 items, newest first
How each company uses it
Anthropic Claude
Anthropic's Claude Opus 5.5 is being discussed as potentially demonstrating a key step towards AGI.
Google Gemini
Google DeepMind has launched a new institute to widen the debate on AGI.
OpenAI ChatGPT
OpenAI is pushing for global AI standards to ensure advanced AI, potentially AGI, benefits everyone safely.
Related terms

Artificial narrow intelligenceAI safetyLarge language modelsAI agentsSuperintelligenceMachine learning

Often confused with
Artificial narrow intelligence
AI designed and trained for one specific task, u
Superintelligence
AI that surpasses human intelligence across virtually all dom
Machine learning
The process of training AI models to learn from data, a founda
Origin

The term 'Artificial General Intelligence' (AGI) was popularized in the AI research community to distinguish hypothetical AI systems with broad cognitive abilities from the specialized 'narrow' AI systems that were becoming prevalent. The concept has roots in early AI research and science fiction.

Try this about 5 minutes

Ask a current AI model to explain a complex scientific concept, then ask it to write a short poem about the same concept. Observe how well it transitions between these different types of tasks and whether it demonstrates understanding beyond simple pattern matching. Try asking an AI to summarize a news article, then immediately ask it to write a fictional story incorporating elements from that article. See if it can fluidly switch contexts and apply its understanding creatively.

Try it: Playbook entries that use thisPlaybook →

The Playbook: reusable how-tos, prompts, configs and tips pulled out of the news.

Questions people ask

Is AGI here yet?
No, AGI is not here yet. While current AI models are becoming very capable, they still lack the broad, adaptable intelligence of humans.
What's the difference between AI and AGI?
AI is a broad field. Current AI systems are typically 'narrow,' excelling at specific tasks. AGI refers to a hypothetical AI with human-like intelligence across a wide range of tasks.
Why is AGI important?
AGI could revolutionize many fields by solving complex problems beyond current human or AI capabilities. However, it also raises significant safety and ethical considerations.

First seen on Subvolts Nov 30, 2021. Updated Oct 16, 2026. Written by an AI model from the items above and reviewed against them weekly; report an error.