AIBloomberg1h ago

some Google employees say Gemini 4 performs well on benchmarks but

Sources: some Google employees say Gemini 4 performs well on benchmarks but struggles with some real-world coding tasks; Google disputes that characterization

TL;DRGoogle's new Gemini 4 excels at benchmarks but stumbles on practical coding work.

Why it matters: Gap between test scores and real-world performance reveals limitations in production AI systems.

Alphabet Inc.'s Google has begun rolling out Gemini 4 Argon, its long-awaited flagship artificial intelligence model …

Read full article

Source: Bloomberg · Opens in new tab