2024: prompt engineer 2025: vibe coder 2026: master of AI agents 2027: hopefully rich enough so it's okay I'm unemployed

Most EdTech products for schools fail because they try to be everything. The winners? Kahoot, Edpuzzle. They solve one problem really well, take 5 minutes to set up, then get out of the way. Teachers don't want transformation. They want their time back.

Hyvä juttu! Tarina oli lyhykäisyydessään se, että koodi on erityislaatuista, kehittäjät ovat pullonkaula ja ohjelmistoyhtiöt, jotka hallitsevat jakelun ja laskutuksen, pärjäävät hyvin tulevaisuudessa. Nyt monien liiketoimintamallit vuotavat ja hinnoitteluvoima heikkene.

Every now and then I need to remind myself: stop building and actually learn how to build better. in vibe coding, it's easy to get pulled into building mode. 2 weeks pass, deep into the project, the momentum feels good. But there's a difference in building fast and building well

startup idea : ads in terminal imagine watching a unskippable 30 secs ad for running "npm install"

The trap with AI vibe coding isn't that it's too hard. It's that it's too easy. You can ship in hours and never ask if you're building the right thing. Speed becomes the goal, talking to users feels like a distraction. That's when you're building wrong.

If English is not your native language, then could you please tell me how old were you when you googled "moat meaning". I can go first: I was pretty old and I did it last week. Year 2026 is the year of moat.

I've been able to locate the secret video that was used as the founding idea for moltbook: youtu.be/qIdq-Bn6trM?si…

Now if we all just loudly announce that it doesn’t anymore make any sense to build anything in software without a real moat. Then doesn’t this lead into a situation where we don’t build anything because those real moats are pretty difficult and a long-lasting moat probably

Are you running a software product business or building a new software product? What is your ”moat”? Let us all know in the replies and we can then think together if that’s a real moat or not.

Figma just closed the last excuse PMs had for not shipping polished UI from AI code. The loop is now complete. Claude Code generates UI. It goes straight into Figma as editable frames. Designers tweak it. Figma MCP sends it back to Claude Code. The entire design-to-engineering

As a product manager, do you know what's the best thing with AI coding agents? When you notice that you as a PM had pushed for something that is not the best way to do something, or is just utterly stupid, and then later need to crawl back to the developers to revert your

Every idea feels taken. Every API already exists. Every SaaS has 12 competitors. So what do we even build now?

Composer 2 is now available in Cursor.

“If your $500K engineer isn’t burning at least $250K in tokens, something is wrong.”

me and Claude