Google is dealing with employee skepticism about the upcoming launch of its next-generation AI model, Gemini 4. Bloomberg reports that as Alphabet’s Google prepares for Gemini 4’s release, some employees question whether the flagship system performs as well as expected in key areas, including coding.
According to the Japan Times, Google has said Gemini 4 achieves top results on several benchmark tests. However, internal criticism persists about how the model performs in practice, particularly for coding tasks, suggesting a gap between benchmark performance and employee perceptions of real-world usefulness.
The two outlets converge on the same core point: Gemini 4 is approaching deployment while internal staff express doubts about aspects of its performance. Where they differ is emphasis—Bloomberg focuses more on the internal skepticism around coding capabilities, while the Japan Times highlights the company’s benchmark claims alongside the internal concerns.
