- Google Deepmind is building an AI email assistant
- The AI will automatically respond to messages in the voice and style of a user.
- The assistant will order and organize entry trays to help reduce the time spent on administration tasks.
There are few more demoralizing modern experiences than waking up, collecting your phone to turn off your alarm and discover an avalanche of emails that bury your entry tray into notifications. Forget about writing again, simply classify them all and determine what need for answers and how quickly you need to answer can eat a whole day. The CEO of Google Deepmind, Demis Hassabis, believes that AI should handle all that organization and response, and he and his team have a plan to embed a secretary of virtual robot in his entry tray so that he never needs to be dragged through irrelevant sales and unnecessary notifications of LinkedIn.
Email is the magical treasure par excellence with an attached dark curse. Miraculous instant communication worldwide is loaded with spam, night demands of immediate action of a boss and the dreaded accidental accidental button. Hassabis believes that AI can restore the good parts of the email handling all “worldly tasks” and recommending answers instead of leaving it full of despair in its count not read of triple digit.
Hassabis told the audience at the recent Sxsw London event that would happily pay enough money to end the current tyranny of the entrance tray. And not in the distant future. Deepmind is working to make the email input trays self -managed, capable of classifying incoming messages, decide which are important and write answers that sound like you. Therefore, the next annoying email not only will not find it well, but will not find it until the AI obtains an answer for approval.
From a marketing perspective, it is a brilliant approach. Many people care correctly about some of the implications of AI could have a weakness for technology if you can clarify even a couple of hours a week for real productive work (or even see your family).
Ia email utopia of AI
Hassabis described his vision of a “universal assistant of AI” that can be proactive and intelligent to act in his name. Instead of regularly giving AI orders and explaining their priorities, AI will learn to observe how it deals with a variety of emails and perhaps a wide description of its approach to different types of messages. He will know not only how he responds to emails, but also which he ignores on purpose and that he is only perplexed when discovering how to respond.
An assistant to that of that caliber, “gives you more time and may protect your attention from other algorithms trying to get your attention,” said Hassabis. “I think we can use in the individual’s service.” “
That is the long -term general panorama. For now, Hassabis said we can expect tools to know when to say: “Thank you for the tracking. I will take a look and get in touch with you soon.” And when to click immediately. All would appreciate an AI that leads us to the zero entrance tray without losing our minds.
You may also like
