Its common to read about LLMs being assessed against standardised human tests, but specialised benchmarks are the gold standard when it comes to evaluating their capabilities. In this first of two articles about LLM benchmarks, we will explore why LLM benchmarks are important, and delve into different types and examples of LLM benchmarks.

They may be small but when it comes to their societal and economic implications, Small Language Models (SLMs) are likely to punch far above their weight. SLMs are cheap and easier to train, set-up, and run than their larger counterparts, and can be designed to excel at specific tasks, making them a perfect tool for businesses and organisations of all sizes.

Back in 2016, DeepMind’s AlphaGo displayed signs of genuine creativity when it defeated one of the world’s best Go players. More recently, OpenAI’s secret Q* model reportedly takes the technology behind AlphaGo forward to enable optimal and efficient decision-making, and could be a hint as to the capabilities of next generation AI models.

Despite the fact that making decisions is a constant in our everyday lives, humans can be surprisingly error-prone when it comes to judgments and making decisions, including and especially the critical ones. Why is this frequently the case and what are some practical steps that we can and should take to minimise such risk?

It’s tough enough trying to get humans to understand each other, much less emotionless machines! Have you then ever wondered how researches have been able to help ChatGPT and other AI models to understand and interpret the complexities and context in human languages, and to respond in a contextually-aware manner? Read on to learn more!

Events frequently take us by surprise, with Covid-19, the war in Ukraine, and the advent of Generative AI, just to name a few. Scenario planning, originally developed as military war games in the 1960s, is today a powerful tool that is used to support decision-makers to envision and plan for substantially different futures.

Most of us are pretty rubbish at assessing uncertain outcomes. This is particular true of situations that are novel, where historical precedent and human experience become irrelevant. But there is hope. SPIES is a simple yet effective framework that can be used by any individual or organisation to improve the quality of their forecasting.

Business model innovation (BMI) does not need to be fancy, or to be based on cutting edge technology and sophisticated data analytics. In fact, innovation can come in the simplest of forms and arise by going back to the basics, as “no frills” Asian barber shop chains such as Japan’s QB House and Singapore’s K-Cuts have done with the men’s haircutting industry.

Digital twin technology is one of several emerging technologies that have the potential to drive step change transformation for the healthcare and life sciences sectors. From improving drug discovery, running in-silico clinical trials, through to optimising hospital operations, and personalising diagnosis and treatment, the possible applications for digital twins are wide-ranging and very significant.