Aaron Levie, the CEO of Box, posted this week that he’s seeing “productivity gains that are an order of magnitude or two above output levels pre-AI.”

An order of magnitude or two. That’s 10x to 100x.

He was reacting to a post from developer Theo Browne about the gap between average and great engineers, and his point is a good one. People who had deep expertise before AI, and who are willing to work differently now, are pulling further ahead. He thinks finance, science and cyber are next.

Then someone in the comments did the math, and I love them for it.

To be 100x more productive, you’d need to finish in one hour what used to take 100 hours. Over an 8-hour day, that’s 800 hours of work. A quarter is roughly 13 weeks of 40 hours, or 520 hours.

So 100x means finishing about 1.5 quarters of work every single day.

Picture your Q1 plan, half of Q2, and a board deck, all done by 6pm Tuesday. Then again Wednesday.

That’s not where most of us are. And I don’t think it’s where we need to be to change how our work feels.

Here’s what 2x looks like in practice.

2x means a day’s work done by lunch. The memo that takes a full afternoon takes an hour and a half. The research you’d normally push to next week happens this week, because it finally fits.

That kind of gain comes from three ordinary things:

1. Picking one workflow you do every week. The meeting prep, the investor update, the pipeline review.
2. Building it once with AI, properly. Give it your context, your template, your examples. Most people skip this part and then decide AI “doesn’t work for this.”
3. Running it again next week and fixing what broke. This is where the compounding starts.

That’s exactly what we do in my AI Leveraged Program. Clients bring a real piece of their work and we build the AI version together, live.

After one of those sessions, a client wrote to me:

“Thank you. This was amazing. I’m obsessed with this.”

That’s what 2x feels like from the inside. You get back hours on a task you used to dread, and you start looking around for the next one.

Levie is right that the ceiling keeps rising, and the people with deep domain knowledge will get the most leverage. The path to 10x runs through a stack of 2x wins, one workflow at a time.

So here’s a question as you start this week: which task, if it took half the time, would change your week?

Start there.

Alex