
In this short talk, the speaker highlights how children today primarily encounter artificial intelligence through their parents' smartphones or tablets, as they are not given personal devices. He argues that most large, globally‑deployed AI models (e.g., ChatGPT) are trained on data that do not reflect the linguistic nuances, cultural values, or heritage of many societies. Using the example of interacting with ChatGPT in Urdu or Hindi, he points out that the model’s responses may miss contextual meanings and cultural references because it was not trained on local values. The speaker warns that if the next generation is educated by such AI, they risk losing connection to their own culture and language. He proposes that each country should develop its own "sovereign AI"—a domestically‑controlled system trained on local data—to ensure AI interactions reinforce, rather than erode, national heritage, language, and societal norms.










