In the rapidly evolving landscape of artificial intelligence, we are witnessing a transformational shift from Large Language Models (LLMs) to Large Reasoning Models (LRMs) to agentic technology leveraging compositionality. This transition signifies a revolutionary change in how businesses, especially in sectors like banking, can achieve enhanced decision-making capabilities through advanced analytical processes.
The evolution of AI reasoning
The journey from LLMs to LRMs is marked by a significant advancement in reasoning. While traditional LLMs excel in language understanding, LRMs leverage Type 2 reasoning during inference to heighten analytical depth without necessitating retraining of existing models. Pioneering systems like GPT-4-03 and Alibaba’s Marco-01 are steppingstones in this evolution, utilising advanced analytical techniques like chain-of-thought reasoning and Monte Carlo tree search to outperform earlier models.
Go deeper with GlobalData
The quadrumvirate: Enhancing AI performance
At the heart of this shift is a redefined framework comprising four elements: compute power, data, algorithms, and crucially, reasoning methods at inference time. This “quadrumvirate” moves beyond LLMs to LRMs and agentic technology to engage in context-driven reasoning, significantly enhancing their performance. By optimising the existing capabilities of trained LLMs through advanced reasoning techniques that include compositionality at inference time using models that have specialised capabilities that are brought together through agentic technology, businesses can harness deeper analytical insights for improved decision-making.
Balancing fast and slow thinking
Cognitive psychology describes two modes of reasoning: Type 1 (fast, intuitive) and Type 2 (slow, analytical). LRMs bridge these modes by dynamically adjusting their approach, allowing for swift actions when necessary while applying thorough analysis in more complex situations. This balance enhances decision-making processes, particularly in sectors like banking, which require agility and depth.
Compositionality: The foundation of agentic technology
The evolution from LLMs to LRMs is further accelerated by the concept of compositionality—enabling systems to deconstruct tasks into manageable components for systematic resolution. Advanced LRM systems utilise techniques like chain-of-thought reasoning and search algorithms, integrating specialised models trained in distinct domains. Unlike classic LLMs, which struggle with compositionality and adaptability, LRMs can effectively navigate complex challenges that extend beyond their initial training data.
Note that this is exactly how IBM’s approach to optimising banking through its proprietary Component Business Modeling (CBM) and industry standards like BIAN operate. These standard industry models break down a bank into its constituent parts thereby enabling specialisation.
How well do you really know your competitors?
Access the most comprehensive Company Profiles on the market, powered by GlobalData. Save hours of research. Gain competitive edge.
Thank you!
Your download email will arrive shortly
Not ready to buy yet? Download a free sample
We are confident about the unique quality of our Company Profiles. However, we want you to make the most beneficial decision for your business, so we offer a free sample that you can download by submitting the below form
By GlobalDataBridging reasoning and compositionality
The interplay between reasoning models and compositionality heralds a new paradigm for AI agents. By building upon the robust foundation of LLMs and infusing advanced reasoning during inference, LRMs achieve heightened intelligence and efficiency. This integration empowers AI to engage in meaningful reasoning, enabling them to tackle intricate tasks in a structured manner by bringing together specialised capabilities – each an expert in its own domain – through agentic technology.
Transforming the banking sector
LRMs coupled with agentic technology and compositionality have profound implications for the banking industry, enabling a dual approach that marries the rapid responsiveness of LLMs with in-depth analytical reasoning. Here are several ways in which generative AI, enhanced with these capabilities, will reshape financial services:
- Fraud Prevention: Real-time behavioural analyses can flag anomalies while understanding broader patterns, positioning banks as proactive defenders against financial crime.
- Credit Scoring: By integrating diverse data sources, LRMs can refine credit assessments and support astute lending decisions.
- Wealth Management: Personalised investment strategies can be generated by analysing market data in conjunction with client behaviour.
- Regulatory Compliance: Real-time analysis ensures adherence to regulations while simulating the impact of potential regulatory changes.
- Customer Insights: Predictive analytics allow banks to anticipate customer needs, enhancing relationship management.
- Operational Efficiency: Workflow optimisation through predictive analysis ensures banks can respond to demands effectively.
Next Steps
- Develop a comprehensive LRM implementation framework
Establish a clear roadmap for transitioning from LLM-based systems to LRMs and agentic technologies. This should include identifying business use cases, aligning with industry standards like BIAN, and integrating compositionality to decompose complex tasks into manageable components. - Enhance AI training and reasoning techniques
Invest in developing and deploying advanced reasoning methods such as chain-of-thought reasoning and Monte Carlo tree search. By optimising inference-time reasoning, businesses can leverage their existing LLM investments while achieving greater analytical depth and flexibility. - Pilot applications in high-impact areas
Launch pilot projects in banking domains such as fraud prevention, credit scoring, and regulatory compliance. These initiatives should demonstrate the practical benefits of LRMs and agentic technologies, showcasing measurable improvements in efficiency, decision-making, and customer experience. - Build scalable, modular AI architectures
Focus on creating scalable architectures that integrate specialised models trained for distinct tasks. Use agentic technology to coordinate these models, enabling adaptive responses to dynamic challenges while maintaining efficiency and accuracy. - Foster collaboration across ecosystems
Collaborate with industry leaders, academic researchers, and technology partners to drive innovation in reasoning models and compositional AI. Establishing shared knowledge and best practices will accelerate the development and adoption of agentic technologies in financial services.
These steps will position financial institutions to capitalise on the transformative potential of LRMs and agentic technologies, ensuring a seamless transition into the era of intelligent reasoning.
A promising future
As we move forward, the transition to LRMs and agentic technology signals a remarkable era in AI capabilities. By emphasising compositionality alongside reasoning, organisations can unlock the potential for deep, meaningful problem-solving. The future promises intelligent agents that not only understand language but also embody sophisticated reasoning processes, leading to smarter, more agile applications across financial services. In this emergent landscape, success will depend on the ability to adapt swiftly and leverage advanced reasoning techniques, marking a departure from traditional operational models. Welcome to the age of intelligent reasoning, where the possibilities are boundless, and innovation thrives.
Shanker Ramamurthy is Global Managing Partner Banking & Financial Markets, IBM Consulting with a focus on bank transformation, core banking and payments. He is also a board member of the Banking Industry Architecture Network (BIAN)
