The world of horse racing poses a challenge due to its unpredictability and various influencing factors. However, the use of artificial intelligence (AI) can potentially enhance the accuracy of predicting winners in this $500 billion market. A new AI tool called HorseGPT has been created to analyze data and make predictions for horse races, including the prestigious Kentucky Derby.
Various methods can be employed to utilize AI in horse racing, such as processing large amounts of data, utilizing pre-trained models, or developing a custom solution. The creators of HorseGPT opted to build a Generative Pretrained Transformer (GPT) AI to analyze data and make predictions. This AI agent, constructed on a large language model similar to ChatGPT by industry leader OpenAI, relies on predefined prompts and training data to inform its actions.
To train the AI model, the creators of HorseGPT began from scratch without any previous horse racing knowledge. They carefully selected prompts to provide the AI with expertise in data science specific to horse racing, including the impact of weather on times, the role of jockeys and riding styles, post positions' significance, and more. They supplemented this with a mix of research papers and blogs on wagering theory, as well as practical knowledge on reading racing forms, interpreting statistics, predictive factors, expert betting strategies, and historical Kentucky Derby data.
When HorseGPT was applied to official racing forms for the Kentucky Derby, it analyzed each race's details based on the provided data to make predictions.
