Standup comedy will be the very last job that AI replaces

Less content, consumed more deeply. Most of what you learn from a book, an essay, or a podcast is not from the information it gives you but from your attempt to make sense of it and use it to solve the questions you are personally trying to answer. You have to fit it into your

The scarce resource in almost every company is synthesis. We are flooded with information: data, customer feedback, missing goals, beating goals, experiments that worked and didn't, competition emerging or changing tactics. The most impactful people in every function can parse

We are entering the age of the IC. Individual contributors will be able to have >10x the impact they've had historically, and this will mean fewer total managers and layers at companies and many more people on senior IC career tracks for the long haul. Three big drivers of this

It's easy to sound smart asking the question "have you also thought about X"? Harder and riskier to ask "are we sure Y actually matters?" or "I think we can cut Z from scope" But in a team full of smart and curious people, usually what they need is the latter - a simplifier

Perhaps the strongest signal that someone will be successful at a startup: they minimize the amount of energy spent worrying about who gets credit or what is fair, and plow it into just doing the work.

When sizing a new initiative (a feature, a marketing channel) we tend to make tidy, reasonable assumptions which add up to tidy, reasonable growth projections. >90% of the time, those growth projects are totally wrong. They are wrong because there is a power law to everything

If you think the customer is acting irrationally, you're wrong. You are just missing some part of their calculus.

One of the highest ROI things you can do: raise the aspirations of those around you

We hear about Satya, Elon, and Jensen 100x as often, but Lisa Su might be the most impressive CEO of the last decade. When she took over as the CEO of AMD in 2014, they had 20 days of cash. They had more than $2B in debt. They had just sold their headquarters and laid off a

never truncate the y-axis

If you don’t think that switching costs are one of the most powerful moats, consider that it took Amazon until 2019 to fully migrate off of Oracle databases. That is 13 years after they launched AWS.