As AI's influence grows, critical questions about its development and impact remain unanswered. Key concerns include the composition of AI training data, with potential inclusion of harmful content and biased perspectives, a secrecy that companies are increasingly reluctant to break despite AI's integration into sensitive sectors. Furthermore, the definition and measurement of Artificial General Intelligence (AGI) lack consensus, despite its use to justify massive industry investments. The article anticipates these transparency and definition issues to be focal points in the coming years, with regulatory bodies like the EU pushing for greater disclosure.