Addition > Multiplication
AI works when it adds capacity to your company. It fails when it multiplies capacity.
The 10x engineer is an old myth, and AI has us hoping it might finally come true. It hasn't. The people chasing 10x as AI-native engineers burn out, and from the companies I've talked to, the real gains are much lower.
A clearer story is emerging, though. Companies that treat AI as adding capacity, rather than multiplying what individuals already do, see higher adoption and happier employees. They also see bigger impact.
Why Multiplication Fails
Multiplication speeds up one person's slice of a workflow, which creates bottlenecks everywhere else. And it leaves the individual with a question: what do I do while the AI works? Most people either multitask or wait. Multitasking might approach 10x in the short run, but we're bad at it, and it causes problems over time. Waiting never gets near 10x and leaves people disengaged from their own work. Neither is good.
Why Addition Works
Addition works because it takes the entire conceive-plan-execute loop off a person, the same way hiring a teammate does. Think of an agent that triages support tickets as they come in, so the team opens their laptops to a sorted queue instead of an inbox. The work gets done faster, and the individual gets time back. Two wins.
Companies that adopt this view fall into a pit of success: people build automations to free up time, which gives them room to think of the next automation.
The Walls
The best AI users are already doing this, even at companies built around multiplication. But they hit walls. Permissions get in the way, or the bot only works for them. Sharing doesn't stick because it doesn't fit the culture. If a company is trying to multiply individual output, adding capacity doesn't make sense.
I think this means we need to stop having individualized tools and start having collective ones. I'm still working out my thoughts here, and I'd love to hear what you think.
