← Back to archive

The real AI value is in apps people actually love

19 May 2025 · 59 views · 1 likes · 0 reposts · 100 blog views

AI Startups automation coding product productivity

We’re in the "there’s an AI for that" era now.

It’s no longer about who has the best model. It’s about who actually builds something people love to use.

Case in point: OpenAI bought Windsurf for $3 billion. At the same time, Cursor is raising money at a $10B valuation after turning down buyout offers. These tools didn’t win because they had some breakthrough model. They won because they fit right into dev workflows and just… worked. Very clean UX with highly specific solutions to user pain points.

You can see it with Perplexity - a tool that started as a basic ChatGPT wrapper. Everyone called it just another UI on top of someone else’s model. But fast-forward a year, and it’s processing 100 million+ queries a week, raising at a $15–18B valuation. Why? Because it's useful and gives you sourced, structured answers fast.

We see the shift in pricing models as well. Instead of flat monthly fees, more AI tools are charging based on usage - how much code you generate, how many tickets you resolve, how many contracts you review. Basically: you pay when the AI actually does something for you.

This isn’t just about dev tools anymore - we’re starting to see vertical AI tools pop up in healthcare, law, education - everywhere. Doctors using AI to write notes. Teachers using AI tutors. Lawyers getting help drafting contracts. And it’s not replacing experts - it’s giving them back time. Making high-quality help way more accessible.

While everyone’s arguing over model benchmarks, the real action is happening one layer up - where people are building apps that actually help.

That’s where the value is shifting. And I think we’re just getting started.

The real AI value is in apps people actually love

Originally posted on X

PS: I'm pretty active on 𝕏 if you'd like to follow along

Related posts

Archive

2026

2025

2024

2018