emerging
Diversified AI Models
The focus is shifting from large language models (LLMs) to a more diverse set of AI models, including smaller, specialized models, multimodal models, and agentic AI.
Themes
Timeframe
near-term
Categories
Impact areas
Detailed Analysis
While LLMs remain valuable for certain applications, enterprises are recognizing the need for a more diverse AI toolkit. Smaller language models (SLMs) trained on curated datasets offer greater efficiency for specific tasks. Multimodal models, capable of processing and generating various data types, expand AI's capabilities beyond text. Agentic AI, focused on task execution, promises to transform work and life by automating discrete actions. "There is no one-size-fits-all approach to AI. There are going to be models of all sizes and purpose-built options-that's one of our key beliefs in AI strategy."
Context Signals
Increased enterprise exploration of LLMs (up to 70% of organizations)
Challenges related to data quality and access for AI training
Emergence of open-source SLMs and multimodal models
Edge
The rise of agentic AI could lead to a future where "there's an agent for that," automating a wide range of tasks and transforming workflows.
Liquid neural networks, with their flexibility and efficiency, could enable AI deployment in edge devices and safety-critical systems.
