- Google is testing a new “light” button in its Gemini application that updates your written application
- The button will produce a clearer and more detailed request without multiple attempts
- The objective is to obtain better gemini responses immediately without writing perfect indications
If you have ever used an AI chatbot, you know that discovering how to express your indications can make a difference to obtain a useful response or gallimatrane. You can spend a lot of time to play with phrases, order of words and level of detail before stumbleding with the correct way to ask a question to AI. Google is testing a new button for your GEMINI assistant to help you reach that ideal message immediately. The next “Light” button, found by Android authorityHe gives his first attempt at a message before sending it to Gemini.
The idea is that, instead of sweating on how to express his warning to Gemini perfectly, he touches this button and lets Gemini Pulle or ‘turn on his initial attempt in something more detailed, more specific and more likely to transmit what he wants to the AI model.
Gemini
This matters a lot when you think about the experience of AI depends on you and your ability to elaborate a notice. It must be specific but not too detailed, thorough but not so much to distract from its main point. Sometimes, you even have to psychoanalyze the AI, discovering strange peculiarities that can affect the result, such as educated or tell the AI not to be lazy.
It has often been useful to ask a chatbot of AI to help her prepare a notice if I am not sure what is the best phrase to convince the information I want from the model. There are also some cases in which AI dates back automatically, but invisibly, it will remode its message before answering. That can be useful, but it could also be the culprit of some of the most erratic answers you have seen.
The ignition button would make it faster to obtain the correct and more transparent application than just make an enamel behind the scene. You write your warning as usual, even if you are only half formed, then press food and let Gemini heat your thoughts scattered in an investigation that is worth sending to AI. Then the improved notice is sent, and voilash, its IA assistant has a much better idea of how to help.
Somehow, this only expands in the suggestions of indications that Gemini sees when opening the chatbot for the first time. Those are much more generic than the ignition button could produce. It would also fit well into the other characteristics that provide variations on Gemini’s production, such as deep research, canvas and image creation with image 3.
The ignition button would be a relatively silent type of update, but that could serve Google’s interests to avoid frustration among Google Gemini users who feel that they cannot make AI satisfy their requests correctly. I could also encourage those who use other AI chatbots who face similar discomforts to visit Gemini and their indications on.
