Second Brain

Over the past 30 days, 99% of my habits have changed.

Instead of logging in to Stripe, my bank, Gmail, and 30 other SaaS tools, I simply got everything I needed through an AI agent connected to all of them via MCPs.

I wake up, open one chat, and say, "Show me yesterday's revenue, any failed payments, and all questions and bug reports from my customers." Done. No 14 tabs. No context switching. No remembering which dashboard has which chart.

The key reason I chose the products I did was their UX. On the database level, all products are pretty much the same, or even if different, they cover the same use cases.

The Traditional UI was basically a layer between me and the database, and now I can use AI Chat to do reads/writes/updates/deletes.

In cases where I wanna see a chart or a table or anything advanced, I ask AI for it, and it either generates it on the fly or I ask it to generate a little tool for this use case. So I ended up vibe coding a super dashboard that works literally with all the tools I use, all in one, only the things I need.

Here's the thing most people don't realize: I only use 5% of any given tool. The rest was just noise to me. So I just picked the 5% from each and put it all into one dashboard.

Think about that. 95% of every SaaS product you're paying for is features you'll never click on. And now you don't have to.

All this made me think: if UX isn't a MOAT anymore, then I don't really care who is the provider of the API...

Whoever gives me a better price for any of the tools, I'll take it.

I'd even want my agent to take care of this: every day it'd check if there are cheaper alternatives to the tools I use, and if there are, it'd cancel the existing tool, download all my data, and transfer it to the new tool.

No migration headaches.

This obviously means that all such tools gonna have their race to the bottom and reach price zero. And I wouldn't even know what tools I'm using. Their brands wouldn't stick in my mind, because I don't care...

I don't care if my email goes through Resend or Postmark or SES. I don't care if my database is Supabase or PlanetScale or Neon. I just care that the thing works and costs the least.

Brand loyalty is gone. Switching costs are gone.

Perhaps the LLM models. Opus 4.6 beats everything else right now, so I'd not switch and I'm willing to pay whatever they ask.

But that game is for the big guys, so it's out of my and your mind as founders.

What's left then?

At first glance... nothing, actually. Tech has no moat. Tech has no defensibility anymore.

Tech business is basically going from 99% profit margins to 1%.

Every SaaS tool is becoming a commodity. Every API is interchangeable. Every database does the same thing. And the AI agent sitting in front of all of them doesn't care about your landing page, your onboarding flow, your clever feature naming, or your freemium funnel.

If you're building a CRUD app with a pretty UI and calling it a SaaS business... yeah, you should be worried. That era is ending.

Think about what I did. I collapsed 30 tools into one chat interface. I built my own command center. I only see the 5% that matters to me from each tool.

Cool. But I'm a technical founder who's been shipping products for years. What about everyone else?

Someone needs to do this for them. The agent layer. The orchestration layer.

I basically built this for myself already. The question is: can you productize it?

I think you can. And I think that's where the next wave of billion-dollar companies comes from.

For most of my vibe coding projects, I'd happily pay someone $199 to purchase the code that I'll then throw into Claude Code and ask to customize it for my needs, and basically, instead of starting from zero, I start with a good default.

The market for such code repositories will be pretty big. The recurring payment will be gone, but selling 10k copies for $199 each every month will equal $2M/mo, and it's pretty achievable (since most users gonna stop paying monthly for their saas, paying once for such repos gonna be an easy decision)

Here's something interesting I noticed. My agent can switch me from Resend to SES in 5 minutes. It can move my database from Supabase to Neon overnight. It can cancel my Stripe account and set me up on a competitor.

But it can't download the collective intelligence of millions of users and bring it with me.

Like... Stripe's fraud detection. It's trained on billions of transactions across millions of businesses. That data IS the product. My agent can't just replicate that by switching to a cheaper payment processor.

If your product genuinely gets better because more people use it... that's the one thing agents can't commoditize. Everything else has no MOAT.

My agent auto-switches my email provider without me knowing. Would I let it auto-switch my bank? My healthcare provider? My insurance underwriter? NO!

The barrier to entry isn't code, it's licenses, compliance, and trust.

I don't wanna pay $49/month for a social media management tool anymore. I used maybe 3 features out of 200. And now my agent does those 3 things anyway.

But what I WOULD pay for: "we'll get you 50 qualified leads this month from LinkedIn." That's not a tool, that's a result. And I don't care how you do it. I don't care what APIs you use behind the scenes. I just care that it works.

The pricing model flips from "pay for access" to "pay for outcomes."

You can actually charge MORE than SaaS. Because you're not competing on features or price. You're competing on "did it work or not."

My agent will optimize my tool spending to zero. But it'll happily pay for results all day long, because results have a clear ROI.

Yeah, tech margins are going to zero. For most SaaS, the game is over.

But every time I've seen something die in tech, the thing that replaces it is 10x bigger. Websites killed newspapers, but created Google. Mobile killed websites, but created the app economy. AI agents are killing SaaS, but they're creating... something we don't even have a name for yet.

So stop building tools. Start building what comes after.