We use third-party cookies in order to personalize your site experience. See our Privacy Policy.

Category thesis

high conviction 27 technologies

Artificial Intelligence

Artificial intelligence has crossed from research curiosity to general-purpose infrastructure in three years – faster than any prior technology adoption. Compute, power, and regulation are the constraints that decide the next adoption tier, not model capability.

The thesis

Category overview

AI adoption has reached 88% of organisations and 53% of the global population in three years. The infrastructure layer (Nvidia GPUs, hyperscaler cloud) captures most value. The application layer remains fragmented. The key tension: exponentially growing demand for AI compute vs linearly scaling chip manufacturing capacity. The EU AI Act (August 2026) will fragment the global market.

Technologies tracked

27 total

AI-accelerated materials discovery AI agents AI code generation AI content authenticity AI for scientific discovery AI infrastructure AI model compression AI model marketplaces AI safety and alignment AI weather and climate prediction Artificial intelligence Autonomous laboratories Computer vision Deep learning Edge AI Embodied AI Federated learning Generative AI Large language models Machine learning Multimodal AI Natural language processing Reinforcement learning Reinforcement learning from human feedback (RLHF) Retrieval-augmented generation (RAG) Speech AI and voice synthesis Synthetic data generation

The rest is inside

The full signal stack, KPIs and catalysts — live.

Every position is backed by talent flows, research and patent velocity, operational signals and a dated catalyst calendar. Log in to CanaryIQ to read the full analysis behind the verdict.