Ghana needs national data cleanup to unlock AI potential –Founder of GradePoint AI

Dr. Gillian Hammah, the founder and CEO of GradePoint AI, has urged the Ghanaian government to refrain from investing in costly attempts to create its own large artificial intelligence models. Instead, she advocated that the country should adapt existing technologies to tackle local issues, citing limited financial resources as a reason to avoid competing with global AI developers.

At the Step Up Executive Breakfast Series, Dr. Hammah emphasized that Ghana should focus on developing smaller, specialized AI models tailored to the country's unique requirements, rather than striving to build a ChatGPT-like system. She argued that such models, known as Small Language Models (SLMs), can be customized using local datasets to address sector-specific challenges.

An example she provided was the Ministry of Health potentially creating an AI model trained on Ghana's health data, while still benefiting from the capabilities and training of larger models. Dr. Hammah stressed that this approach would allow Ghana to leverage AI advances without the monumental financial commitment required to build entirely new language models.

She urged the nation to concentrate on practical applications in healthcare, agriculture, and food production, stating that the pursuit of a ChatGPT-like model is a waste of time. Additionally, she highlighted poor data management as a significant hurdle to Ghana's AI aspirations, advocating for a nationwide effort to digitize and standardize government records.

This reliable and accessible data, she believes, would lay the groundwork for developing specialized AI applications that cater to the country's specific needs.