The Minister for Communications, Digital Technology and Innovations, Sam Nartey George, is urging academics to build Ghana’s own Artificial Intelligence language models to reduce reliance on foreign-built systems.
The Minister says AI models built in Europe or America will not work for Ghana because they cannot capture the true meaning of Ghanaian languages.
Speaking at KNUST during a tour of the One Million Coders labs in the Ashanti Region, he said the government has decided that Ghana’s national language models must be built locally.
“One of the key things that AI will do, if we are not careful, is that it will create an extinction of our indigenous African languages. And it will create a big divide between the people and us,” he warned.
He tasked experts at KNUST to lead the development of a Twi language model due to its location within Asanteman, while other universities will lead models for other Ghanaian languages.
According to him, this is the only way to prevent the extinction of indigenous languages and ensure AI benefits all Ghanaians.
The Minister explained that in many Ghanaian languages, the same word can have different meanings depending on tone.
He cited Ga and Dangbe as examples, where a word like “nu” can mean man, water, or drink depending on how it is pronounced, something foreign developers cannot interpret correctly.
“When you take, for example, my local dialect of Ga or Dangbe, the same word written can have multiple meanings. So if you take a word like ‘nu’, it can mean man, it can mean water, it can mean drink depending on the inflection,” he explained.
His solution is to pair academic knowledge with traditional authority.
He said government has identified academic champions – lecturers and professors – in various universities to lead the drive, working closely with traditional leaders who are custodians of the languages.
Another key solution is the creation of voice interfaces in local languages.
The Minister explained that while AI solutions for agriculture like precision farming, weather sensing and crop disease detection already exist globally, they remain inaccessible to many Ghanaian farmers if they only speak English.
He said the government wants voice-enabled AI that farmers can interact with in Twi, Ga, Ewe, Dagbani and other local languages, and plans to use the same technology, combined with the newly deployed 5G spectrum, to deliver services like telemedicine to remote communities.