James' World 2

To find out if technology could do a better job than I could at making my schedule work for me, I tested out a free trial of Blockit, a new AI-powered agent that integrates with a user’s calendars and email. When signing up for the tool, Blockit promised me that in as little as five minutes, it could learn the same amount of information about my schedule, habits, and preferences as a human EA might over the course of several months.

Here’s how Blockit works: The AI agent learns your preferences for taking meetings, including when and where you like to conduct certain kinds of business. Then, you can copy the Blockit bot into emails or Slack messages with your contacts and give it instructions to set up a meeting at your chosen time and place. 

It sounded fantastically simple, but after using the tool, I realized that letting Blockit’s AI into my schedule required more than a little work on my part, too. Here are my three biggest takeaways from letting AI into my schedule for a week. 

You need to work to make it work for you

Blockit’s onboarding process involves answering multiple questions about your habits and schedule, some of which got me thinking a little more about where, in fact, I like to work. So if you like to take certain meetings in a coffee shop near your office, you need to tell Blockit the exact address, and the AI will make a note of it for future reference. Similarly, if you have an office or work from home on certain days, Blockit will log that, too. 

Doing this means that when you copy Blockit’s bot into an email with a contact that you want to get a coffee with, the bot will schedule a meeting at your preferred spot, invite the other person to it, and block off the time on your calendar that it will take you to get there from wherever you told it you would be working that day. That’s extremely helpful! 

But it also requires you to make some concrete decisions about where and when you will be working—and that’s not always totally obvious if you are in an industry that regularly puts you in many different locations on short notice. 

Blockit, to its credit, can keep up—it will even ask you to confirm if you are traveling if you tell it to set a meeting in an unfamiliar city. But if you are a busy CEO, keeping your AI agent up to date on your schedule might not always be top of mind. 

Another interesting Blockit feature is its codewords function. Users can teach the AI codewords that trigger certain actions: For example, say I sign off an email agreeing to a meeting with “best wishes” and copy Blockit to set something up. I could have already set “best wishes” as a codeword, meaning that this meeting is not high priority, can be set sometime three or four weeks away, and can be canceled if I get another, higher priority request for the same time between now and then. 

It’s a clever idea, but again, I had to go through the work of teaching Blockit my codewords, a process that the desktop app doesn’t make particularly intuitive.