Google has launched a new feature for its Gemini 3 model called Deep Think. This mode is designed to tackle complex problems that require extensive and detailed analysis. Its goal is to emulate technical and structured reasoning processes, useful in fields like engineering, design, or research, going beyond immediate responses.
Architecture and application in technical processes ⚙️
The Deep Think mode operates by breaking down queries into sequential logical steps, generating internal chains of thought before providing a final response. This allows for a deeper scrutiny of premises, consideration of multiple variables, and structuring solutions step by step. The technology aims to be a support tool in tasks that require detailed planning or problem-solving with many interconnected factors.
Does this mean it won't say "as a language model, I can't..." anymore? 😏
Hopefully, this thinking mode will make the AI at least simulate having read the question twice before responding. Maybe now, when asked to design a flange, instead of a generic sketch, it will give the steps and then add: Step 11: Look in the drawer for the wrench that your coworker surely didn't return. A technical advancement, but workshop common sense remains human.