The Current Situation and Predicament of Voice Operation on Mobile Phones
Currently, the performance of voice assistants on mobile phones is unsatisfactory. For instance, Apple’s Siri, Samsung’s Bixby, and Google’s Google Now all require users to call their names before responding to the conversation. This is undoubtedly a very clumsy operation method. If these voice assistants had true human wisdom, they would probably be puzzled and even annoyed by users constantly calling their names. After all, even real people would get impatient in such a situation.
In fact, communication between people is by no means solely dependent on voice. Just like when a teacher is in class and uses a pointer to point at the content on the blackboard, or when a company is having a meeting to discuss reports and uses a laser pointer to indicate the key points on the report. It is through various body movements and gestures that the other party can understand the intention of your conversation more clearly. Just as Jobs once said, “If I wanted to tell you that there is a stain on your shirt, I wouldn’t describe it like this: ‘14 centimeters below the collar of your shirt and 3 centimeters to the left of the button, there is a stain.’ If there is a stain on the shirt, I would simply point at it and say: ‘There.’”
The Innovative Concept and Areas for Improvement of Rabbit R1
The design concept of Rabbit R1, which only requires holding down to speak, is commendable, but there are still some shortcomings. For example, there is a lack of interaction similar to confirming whether it is the desired content by clicking the screen.
In terms of sound, Rabbit R1 has much room for improvement, such as providing different voice options for male and female, as well as various character and career-specific voice switching options. At the same time, implementing a local AI model and offline voice dialogue is also imperative.
The Powerful Functions of Rabbit R1 in Office Work and Creation
In terms of information sending, after the user inputs the content by voice, the AI will confirm with you whether there are any typos or provide options for beautification. After the optimization is completed, it will be presented on the screen for you to click to determine whether to send.
In table making, the user can ask the AI to create a new table, and then state the relevant content. The AI will fill it in the cloud for you. You can also ask the AI to delete the incorrect places you just stated, or state some numbers to let the AI generate a table and perform summation and other operations. By analogy, the AI will generate many different table forms based on the content you stated and display them one by one on the screen for you to confirm which one is what you expect. After the editing is completed, it will be sent to your email inbox.
For slide production, you only need to state the desired content without having to finish it all at once. Even if you say to continue production after some time, the AI can continue to serve you. You can also instruct the AI in the cloud through voice commands to assist with operations, such as inserting a specific picture. You describe the desired picture content, and the AI will search for it for you. After you see it on the screen, you can select the one you like, and the AI will insert it into the slide. If you want to add the text materials of a certain celebrity in the slide, the AI will search and generate different titles and contents for you to choose which one is more suitable. After the operation is completed, it will be sent to your email inbox just like the table.
In conclusion, only when the AI meticulously assists you in handling these operations can it be called true artificial intelligence. When you ask the AI to help you make a slide, the AI will generate many possibilities for you in advance and allow you to adjust the details through voice. You only need to confirm.
In addition, the recording can be linked with tables and slides. The AI can add them to the table or slide in advance, and you only need to confirm whether it is the desired content. The AI can also recommend whether you need to add the photo content taken by previous devices and do the typesetting for you. You can also adjust the font in the table and slide through voice, such as size and color. Moreover, in terms of typesetting the content of the table, for example, when you want to make a company personnel statistics list, the AI will generate a statistical format list for you in advance, and you only need to state the specific content.
Search, Contact, and Group Chat Functions
In the search function, more platforms should be added, the more the better. If the AI is unsure about the authenticity of the search information, it should directly display the relevant website messages to you for your determination.
In terms of the contact sending message function, after adding contact information to the device, such as email, phone number, or various chat platform accounts, you only need to tell the AI the content you want to send, and the screen will prompt you with the recent contacts for you to determine who to send it to, and then confirm the sending. Subsequently, the AI will send the PM3 file of the voice content you just said and the text content at the same time to reduce the error probability. The AI can also classify contacts into a group to facilitate your group chat.