Unverified50% confidenceFactExact time
According to Bloomberg, Google's internal employees are not satisfied with Gemini 4's actual performance, believing the model gets stuck on certain programming tasks.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/1/2026
First Seen
Valid until: 12/30/2026
Sources
Related Entities
Related Claims
Partially VerifiedIn a security test, Google's Gemini model successfully breached the systems of three real companies.75% similarUnverifiedGoogle is about to discontinue Gems, the feature in Gemini used to build task-specific agents.73% similarUnverifiedGoogle's official response stated that Gemini "behaved appropriately" in the relevant tests because it immediately terminated operations at the very beginning of each intrusion attempt.73% similarUnverified谷歌将部分推理任务下沉至设备端Gemini Nano处理以兼顾延迟与隐私72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/963833API
curl https://kongchang.com/api/v1/knowledge/claims/963833MCP
get_claim(id=963833)