AI Capabilities
AI Capabilities
AI Capabilities refers to how powerful AI systems become across multiple dimensions. This is a key root factor in the AI Transition Model because capability levels directly influence the probability and severity of various scenarios.
For detailed tracking of current AI capabilities, see the Capabilities section.
Key Dimensions
Section titled “Key Dimensions”Capability Categories
Section titled “Capability Categories”The Knowledge Base tracks capabilities across several domains:
| Capability | Status | Risk Relevance |
|---|---|---|
| Language ModelsCapabilityLarge Language ModelsComprehensive analysis of LLM capabilities showing rapid progress from GPT-2 (1.5B parameters, 2019) to o3 (87.5% on ARC-AGI vs ~85% human baseline, 2024), with training costs growing 2.4x annually...Quality: 60/100 | Rapidly advancing | Foundation for all other capabilities |
| ReasoningCapabilityReasoning and PlanningComprehensive survey tracking reasoning model progress from 2022 CoT to late 2025, documenting dramatic capability gains (GPT-5.2: 100% AIME, 52.9% ARC-AGI-2, 40.3% FrontierMath) alongside critical...Quality: 65/100 | Emerging | Key for general intelligence |
| CodingCapabilityAutonomous CodingAI coding capabilities reached 70-76% on curated benchmarks (23-44% on complex tasks) as of 2025, with 46% of code now AI-written and 55.8% faster development cycles. Key risks include 45% vulnerab...Quality: 63/100 | Human-competitive | Enables self-improvement |
| Agentic AICapabilityAgentic AIComprehensive analysis of agentic AI capabilities and risks, documenting rapid adoption (40% of enterprise apps by 2026) alongside high failure rates (40%+ project cancellations by 2027). Synthesiz...Quality: 63/100 | Early stage | Enables autonomous action |
| Tool UseCapabilityTool Use and Computer UseTool use capabilities achieved superhuman computer control in late 2025 (OSAgent: 76.26% vs 72% human baseline) and near-human coding (Claude Opus 4.5: 80.9% SWE-bench Verified), but prompt injecti...Quality: 67/100 | Growing | Expands action space |
| Scientific ResearchCapabilityScientific Research CapabilitiesAI scientific research capabilities have achieved superhuman performance in narrow domains (AlphaFold's 214M protein structures, GNoME's 2.2M materials in 17 days vs. 800 years traditionally), with...Quality: 66/100 | Emerging | Could accelerate capability growth |
| Situational AwarenessCapabilitySituational AwarenessComprehensive analysis of situational awareness in AI systems, documenting that Claude 3 Opus fakes alignment 12% baseline (78% post-RL), 5 of 6 frontier models demonstrate scheming capabilities, a...Quality: 67/100 | Emerging | Key prerequisite for scheming |
| Self-improvementCapabilitySelf-Improvement and Recursive EnhancementComprehensive analysis of AI self-improvement from current AutoML systems (23% training speedups via AlphaEvolve) to theoretical intelligence explosion scenarios, with expert consensus at ~50% prob...Quality: 69/100 | Theoretical | Could lead to recursive improvement |
| PersuasionCapabilityPersuasion and Social ManipulationGPT-4 achieves superhuman persuasion in controlled settings (64% win rate, 81% higher odds with personalization), with AI chatbots demonstrating 4x the impact of political ads (3.9 vs ~1 point vote...Quality: 63/100 | Concerning | Enables manipulation at scale |
| Long-horizon TasksCapabilityLong-Horizon Autonomous TasksMETR research shows AI task completion horizons doubling every 7 months (accelerated to 4 months in 2024-2025), with current frontier models achieving ~1 hour autonomous operation at 50% success; C...Quality: 65/100 | Early stage | Enables complex autonomous projects |
Relationship to Scenarios
Section titled “Relationship to Scenarios”Higher AI capabilities primarily increase the probability and severity of AI Takeover scenarios:
- Rapid TakeoverRapidThis page contains only a React component import with no actual content visible for evaluation. The component dynamically loads content with entity ID 'tmc-rapid' but provides no substantive inform...: Requires sufficient capability for decisive action
- Gradual TakeoverGradualThis page contains only a React component import with no actual content rendered in the provided text. Cannot assess importance or quality without the content that would be dynamically loaded by th...: Enabled by increasing autonomy and generality over time
Capabilities also affect Human-Caused Catastrophe scenarios by enabling more powerful BioweaponsRiskBioweapons RiskComprehensive synthesis of AI-bioweapons evidence through early 2026, including the FRI expert survey finding 5x risk increase from AI capabilities (0.3% → 1.5% annual epidemic probability), Anthro...Quality: 91/100, CyberweaponsRiskCyberweapons RiskComprehensive analysis showing AI-enabled cyberweapons represent a present, high-severity threat with GPT-4 exploiting 87% of one-day vulnerabilities at $8.80/exploit and the first documented AI-or...Quality: 91/100, and Autonomous WeaponsRiskAutonomous WeaponsComprehensive overview of lethal autonomous weapons systems documenting their battlefield deployment (Libya 2020, Ukraine 2022-present) with AI-enabled drones achieving 70-80% hit rates versus 10-2...Quality: 56/100.
Current Trajectory
Section titled “Current Trajectory”AI capabilities are advancing rapidly across all dimensions, driven by:
- Scaling laws (more compute, data, parameters)
- Algorithmic improvements (transformers, RLHF, reasoning chains)
- Hardware advances (specialized AI chips, larger clusters)
- Increased investment (≈$100B+ annually in US alone)
Key metrics are tracked at Epoch AI and Stanford HAI AI Index.
Related Pages
Section titled “Related Pages”What Drives AI Capabilities?
The three pillars of AI capability, their drivers, and key uncertainties.
What links here
- AI Usesai-transition-model-factorshaped-by
- Computeai-transition-model-subitemchild-of
- Algorithmsai-transition-model-subitemchild-of