AI: One Mind, Many Hands
On how AI, the rise of agents and parallelisation will change knowledge work forever
One of the more fascinating shifts happening with AI right now is the rise of this new kind of “parallelisation” in how we work.
Here’s a real-life example that just happened a few minutes ago, that prompted me to write this post.
Someone on our marketing team sent me a list of 36 industry people they recommended I follow on LinkedIn.
I was mid-task, but I wanted to get this done quickly, so I just opened up Composite, dropped in the list of LinkedIn URLs, and asked it to take over a browser window and start following each person one by one.
Then I just… went back to work in another window. Five, maybe ten minutes later, it was done.
Now, on the surface this seems like a small, even trivial, thing. Pretty basic browser automation.
But I think it’s a glimpse of what’s coming.
Today, we’re still largely constrained by the fact that we can only interact with one thing at a time. One browser window. We can only point, click and type in one place, at a time.
But computer-control-AI tools like Composite (as well as everything else in AI, agents and LLMs) hint at a future where you’re working on 3–5 things at the same time.
You could be writing a doc, updating a spreadsheet, drafting emails, all happening in parallel, with AI acting as a kind of executive assistant that actually executes the work.
This opens up some really interesting questions:
Do we retain less knowledge when we no longer engage in the friction of doing? If AI handles the works, are we learning, or just moving forward?
What does “focus” even mean in a world where tasks run in parallel without us? Can we still concentrate or get into deep work flow, when there’s always something else “working” in the background?
What happens to craftsmanship when speed becomes the goal? Is there still room for “slow work” in an AI-accelerated world?
Needless to say, these are interesting times.
