Taku 2.0 Beta: Built With Users, Not Just for Them

Tomorrow we ship the biggest change to Taku since we started.

We're calling it Beta. Not as a disclaimer. As an invitation.

Why are we calling it Beta?

The boring answer is that it isn't finished.

True, and not interesting. Many products wait until everything feels polished before inviting users in.

The real answer is a decision about who gets to shape what comes next.

The next Taku should be built with its users, not just for them.

What you can do in Taku 2.0 Beta

Discover. Find AI apps, agents, skills, and workflows created by the community.

Run. Use them instantly without dealing with setup, dependencies, or deployment.

Remix. Customize existing setups with your own tools, data, and goals.

Create. Build your own AI workflows and turn ideas into reusable systems.

Share. Publish your setup as a Stax for others to discover and remix.

For, versus with

Building for people is the normal way, and it isn't cynical. You research, you decide, you build in private, you launch. Feedback arrives after the decisions have hardened. By then the only vote left is whether to stay.

Building with people means getting you in before the concrete sets. While the shape can still move. While "actually, this is backwards" is a cheap sentence to hear instead of an expensive one.

Beta is what that looks like on a calendar. A stretch of time where the product is real enough to use and soft enough to change.

The catch is that it only works if the people using it treat it that way. Which brings us to the ask.

What we're asking of the first users

If you join the Beta, you're not just testing software. You're becoming one of the first builders of Taku.

We want you to:

• find workflows you actually need
• create setups we never imagined
• tell us where the experience breaks
• show us what should exist next.

We want the halfway. We want the moment you closed the window. That moment is invisible to us and obvious to you, and it's the single most valuable thing you can hand over.

What we're building toward

Here's the shape of 2.0, without the full tour:

Discover → Run → Remix → Create → Share

Discover. Find what other people have already built instead of starting at a blank page.

Run. Use it immediately, with no setup standing between you and the thing.

Remix. Make it yours. Your context, your data, your goal.

Create. When the thing you want doesn't exist, start from the idea.

Share. Send it back out so it becomes someone else's starting point.

That's a loop, not a funnel. Each step feeds the one after it, and the last step feeds the first.

I'm deliberately not writing the full product tour here. That's tomorrow.

What Beta actually means, practically

Things will break. Some of it will be embarrassing.

We would rather be caught early than be polished and wrong. A rough thing you can fix beats a finished thing nobody needed, and the gap between those two outcomes is decided in the first few weeks of real use.

If that trade sounds unappealing, wait for the version after this one. Genuinely, no hard feelings. If it sounds like the interesting part, you're the person we're building this with.

One more thing

Every article this week has been about a gap.

AI that could think but not execute. Tools you found and never managed to use. Things you built that nobody ever found.

Calling this Beta is us admitting we don't get to close those gaps alone.

Taku 2.0 Beta launches Aug 15. If you've ever saved an AI tool and never used it, this is for you.

Come build with us.

Borrow brilliance. Make it yours.