This year, HubSpot, like seemingly all tech companies, has incorporated AI into its product line.

Don't feel like writing an email subject? Write the email, and AI will make a subject line for you.

Want to make a report but don't know how to do it? Write out an AI prompt, and it'll make a HubSpot report for you.

Having trouble describing your automations? Make the workflow and AI can write a description for you.

It all sounds good. But my mental image of AI being a somewhat useless intern still seems to explain most of its utility: it gives you something, which is not nothing, but you almost always need to heavily edit--often to the point of completely changing--what it drafts. It's like the intern that can't be trusted to actually do anything.

At this point, I dropped this message into our subject line generator. Here's how AI thinks this email could appear in your inbox:

That's ... fine. But I'm going to need to re-do it to make it usable.

This lack of real utility makes me think AI, in its current "use cases," could be like the smart speaker bubble from a few years back. Cool technology, but wildly overblown in it's potential impact. Voice control didn't become the new UI and smart listening didn't become the new I/O: both became a nice little niche in how some people like to use in their devices. I liked this analogy so much, that I stole it from Ben Evans, whose essay is the first link below.

The second link for your attention this week is a really nice take-down of the AI executive order. I'm not sure I agree with all of it, but it's an excellent read and a fair critique of both this executive order and the executive order as a way of governing.

The AI subject line generator did a bit better at this point in the email. "AI: Overhyped or Overrated?" to which: why not both

Enjoy the reading!

Reading