One idea, one quote, one question. 90 seconds.
The Idea
In the spring and summer of 1957, J.C.R. Licklider made himself the subject of an informal time-and-motion study. He tracked what he actually did during the hours he considered thinking. About 85% of that time was spent getting into position to think: searching for information, plotting graphs, calculating, and converting data into comparable form. Hours went into the plotting alone. When the graphs were done, the relationships were obvious at once, but the plotting had to be done in order to make them so.
Three years later, he published a paper called “Man-Computer Symbiosis” in which he laid out the division of labor that we’re still trying to get right. His argument was that the human should be the one deciding what to work on, what questions to ask, and whether the result is any good. The computer should handle all the clerical and mechanical work between those calls. Not replace the thinking, but clear the path so the thinking can actually happen.
Nirav Tolia described a modern version of this on the podcast this week. Nirav is the co-founder and CEO of Nextdoor, which grew to more than 100 million verified users. At one point he shared a heuristic that’s been going around Silicon Valley called 20-60-20, and he used podcast prep as the example. The human does the first 20% to understand who the guest is, why they belong on the show, and what matters. AI handles the middle 60% by expanding themes and generating possible questions. Then the human comes back for the last 20% to choose, order, refine, and add the judgment that turns raw material into something worth saying.
Nirav didn’t claim to invent it, and it’s a heuristic, not a formula. But I think it’s useful because it gives AI a large role without giving it the final word. The first 20% is where you decide what matters, and the last 20% is where you decide what’s good enough. Skip either one and you end up with something that looks finished but isn’t pointed at the right problem. AI can challenge both of those decisions, but you still own them.
Quote
“The essence of strategy is choosing what not to do.” Michael Porter
Question
What is one piece of work this week you should make AI challenge instead of making it do for you?
Listen
Nirav Tolia: The Man Who Connected 100 Million Neighbors Says AI Is About to Make Us All Lonelier. The 20-60-20 idea is what I built this newsletter around, but most of the conversation is about how Nextdoor created trust through identity verification, why Nirav thinks AI will make genuine local community more valuable, and what it took to come back as CEO six years after stepping away and lead a turnaround to profitability.
Read
Co-Intelligence by Ethan Mollick. Mollick is a Wharton professor who treats AI as a co-worker, co-teacher, and coach. It is a practical guide to learning where the tool helps, where it misleads, and how to work with it without confusing a confident answer for sound judgment.
If you love this content (please share it), but also check out my Podcast, and connect with me on YouTube / Twitter.
— Scott