When you've spent $200 on tokens building an app nobody asked for

Your strategy must exclude.

We'll have to read strategy docs differently now that everyone uses AI. A well-structured doc used to suggest someone thought hard about the problem. Now it might just mean someone wrote a (decent?) prompt.

For 20 years, good product management meant saying no to "should have" feature requests. Every extra feature added maintenance, testing, support, complexity, etc. But if agents make most of that cheap, maybe feature bloat stops being a problem. A product could (will?) have thousands of niche, customer-specific capabilities. You just wouldn't see most of them. – at Sydney, New South Wales

Bring back the ability to open "New Chat" in a new tab on web. Cmd + click doesn't work anymore...my muscle memory is deeply offended.

The below is not mine but I really like it. The more I think about it as a parent, the deeper it feels. --- Someone asked a grandmother: "What is the most important thing you can give your children?" I expected her to say: love, education, a family. She said: "A place they won't be afraid to come back to when they've fucked everything up."

Shitposting from Fiji

One of the nicest things a PM can hear from engineering is: "There's a much simpler way to do this."

Good founders change their minds. The useful bit is saying what changed and why so the team doesn't have to guess.

One of the best things a PM can do for engineering is give them the problem properly and leave some room to solve it. They usually know things you don't.

There's usually a point where a product team has too much work in progress and everyone knows it. Nobody wants to be the person who suggests stopping their project though. So engineering and/or product gets blamed for being slow.

One of the nicest changes in a growing company is when product decisions stop depending on whether the CEO/founder happens to be in the room.

There's something really healthy about a team that can ask: "What would we stop if we did this?" before getting excited about the new thing.

A surprising amount of product work starts with: "While we're in there, could we also ..."

One of the best signs in a product team is when someone can say: "We're not doing this" and nobody treats it like failure.

"Business critical" is one of those phrases that stops meaning anything once 10 things in the same meeting are business critical.

"A customer asked for it" is probably one of the weakest reasons to build something and one of the hardest arguments to push back on.

We spent years trying to increase engineering capacity. Now AI is doing it and we're discovering that a surprising amount of company processes were designed around engineering being slow.

This operating model is (surprisingly) still common for all kinds of projects even small, low-risk ones: Team has a decision to make → ask CEO → wait → CEO replies → continue.

It might be just me but DHH seems to be moving at an insane speed right now. Not just on the coding front. The whole Omarchy thing seems to be moving from idea → product → community → ecosystem ridiculously fast. Meanwhile it took me half a week to build an outbid.lol clone 😅 I know comparing yourself to DHH is probably not the healthiest benchmark but X lately has completely warped my sense of what "fast" looks like. Anyone else feeling strangely slow compared with what you're seeing people ship lately?