Google grapples with employee skepticism about new Gemini 4
The company said Gemini 4 posted leading scores on several benchmark tests, while internally some criticize how it performs in areas such as coding.
Google, the tech giant under Alphabet Inc., has recently unveiled Gemini 4 Argon, its cutting-edge artificial intelligence model. This new model, revealed on Wednesday to a select group of cybersecurity partners, is set to be expanded to paid subscribers following further testing.
According to Google, Gemini 4 has achieved top scores on several benchmark tests, including outperforming OpenAI's Astra model in one area that evaluates security skills. However, internal sources suggest that the model's real-world performance is less impressive. Employees handling actual tasks report that Gemini 4 struggles with certain coding tasks, a claim corroborated by individuals with insider knowledge.
Despite posting leading scores on various benchmarks, the skepticism within Google underscores the challenges of translating theoretical performance to practical application. The discrepancy between benchmark performance and real-world effectiveness highlights the complexities of developing advanced AI models.
Written by urgent.news from Japan Times's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
