Joo Tat

Building

[Joo Tat @heyyyjoo]

Was planning out my work for the year when this struck me.

To do big things is to do things that have delayed but outsized gratification. And to not burn out while doing things that have delayed gratifications is to do project management well with your past and future selves.

[Visa is doing marketing consults (see pinned!) @visakanv]

This is where people start to fall off

8. Project management – what substantial things do you wanna do, or think you wanna do? How do you design the tasks that comprise it? how do you communicate and collaborate with future versions of yourself to make progress on these tasks?

[Joo Tat @heyyyjoo]

@soontiago Magical internet money is the best money

[Joo Tat @heyyyjoo]

@soontiago Niceeeee!!!!

[Joo Tat @heyyyjoo]

@NickADobos Composer 2.5 has been great. Can’t wait to see what comes next

[Joo Tat @heyyyjoo]

@israfill Gotta be careful

[Joo Tat @heyyyjoo]

My heart goes out to Max. Nothing worse than spending time to build, only to launch to crickets.

Always advisable to figure out how to establish a pipeline of potential customers BEFORE launching.

You CAN “launch” first then worry abt customers later. But expect crickets, otherwise you’re setting yourself up for disappointment.

[Max Howell @mxcl]

I think I will give up. I have made a bunch of very interesting things over the last 6 months. But I cannot get any attention for them and don’t know how. My son is 4 and he deserves my time. I'm going to go and get a normal job. My time making things is over I think.

[Joo Tat @heyyyjoo]

Maybe it’s because good writing is interesting.

Interesting is by definition, out-of-distribution.

You can still wrangle models to produce good writing tho.

Same way you need to wrangle models to produce (actually) good software.

[Sully @SullyOmarr]

we basically have a digital coding god

but literally not a single model can write even if its life depended on it

what happened? just too much garbage text on the internet? no one cares about writing?

[Joo Tat @heyyyjoo]

@ramsri_goutham You mean like robot vacuums? If you trust Reddit you can check this out https://t.co/KUvJ9ncWfG

[Joo Tat @heyyyjoo]

Here’s my main first impression of Sauna: I don’t see a clear winning use case yet.

At least from my perspective, Sauna currently feels like a broad AI layer on top of the apps you already use. It can suggest what to do, help find information, and has a multiplayer/collaboration angle around understanding what other people are doing. But I don’t yet see the specific use case where Sauna is clearly much better than existing alternatives.

For an early product, I think it would be useful to have a sharper wedge: a specific group of people, in a specific situation, with a painful problem where existing solutions are poor, and where Sauna is obviously the better answer. Maybe that wedge already exists, but as an outsider looking at the website and demo, it is not immediately clear to me.

This feels especially important because Sauna asks users to overcome a meaningful amount of friction and anxiety. To unlock the value, users may need to connect sensitive apps like email, Slack, and Notion. If the multiplayer value is important, they may also need to convince teammates to connect their own sensitive apps. That creates a big trust and coordination hurdle, so the value proposition needs to be extremely clear before people will make that jump.

One analogy I think about is Notion. Notion is now a very broad horizontal product: people use it as a CRM, Jira alternative, team wiki, notes app, etc. But early on, I believe it had a much simpler starting point: document and knowledge organization. The product and communication was focused on a better way to store, structure, and share notes and docs compared to alternatives like Google Drive, Dropbox, or scattered documents. People could use it for their own notes and documents first. Then, when they eventually shared a page with colleagues, those colleagues could immediately see the value because the page was easy to navigate, clear, flexible, and beautiful.

I wonder what the wedge could be for Sauna. I noticed that the Solutions page seems to heavily feature Sauna in Slack, as an assistant that has access to shared context. Is that something that has been resonating better with users?

One possible wedge could be someone who is overwhelmed by Slack because they have too many messages and threads to respond to. They could drop their personal assistant into a channel to help reply on their behalf, using context shared with Sauna, and escalate when it is unable to answer confidently. That might also create an easier mental model around access: the personal assistant in Sauna has access to more private context, while the team-facing assistant in Slack has more limited, scoped access.

From there, if colleagues see the assistant working and want their own, that could be a natural path into the multiplayer or “git main branch” idea. Individual assistants could start to merge shared context and provide better help, suggestions, and coordination over time.

I’m not sure if this is the right wedge. The answer may already be visible in the product’s usage patterns: who is sticky, what they are using Sauna for, and where they are getting repeated value. But I think the key question is: what is the specific initial use case where Sauna is not just useful, but dramatically better than the alternatives? Once that is clear, I think the product / website / demo should make that use case extremely obvious to the users who need it.

(Btw I'm speaking with Ryan tmr regarding the PM role. Which was what led me to explore Sauna as part of my own research. Thought I might as well share my first impressions here)

[Joo Tat @heyyyjoo]

@SullyOmarr You can wrangle a model to produce good writing tho. Like how you can wrangle a model to produce good software.

[Joo Tat @heyyyjoo]

@SullyOmarr Good writing is interesting. Interesting is by definition out of distribution.

[Joo Tat @heyyyjoo]

@nurijanian This is a banger for the number of engagements it got

[Joo Tat @heyyyjoo]

@0xwhrrari Bookmarked and saved to my home scren

[Joo Tat @heyyyjoo]

@businessbarista It’s easy to get lost in building when building gets easier but isn’t the bottleneck. Saving this to home screen to remind myself from time to time.

[Joo Tat @heyyyjoo]

@NickADobos What are you building?

[Joo Tat @heyyyjoo]

@Replit Maybe when the system detects that a way of working with the agents is inefficient (e.g. user doesn't have the vocab to precisely describe a change), the system can intervene and introduce some concepts that could help.

Same principles as this: https://t.co/szhx2ErdLo

[Joo (Coding Proj) @0to1000_pdEng]

Student describes what he/she wants to build. AI tailors a coding course that builds towards what the students wants to build.

[Joo Tat @heyyyjoo]

I built https://t.co/l8ZGakcl1B on @Replit 2 years ago, then eventually outgrew it as the app got more complex.

Recently, I jumped back in to vibe-code a new idea.

First surprise: the file panel is hidden by default. Took me a few minutes just to find it 😅

That’s interesting because even when I’m vibe-coding, I still look at the file structure to build a mental model of the app.

It helps me better direct AI agents when improving or debugging things.

It also helps me know what I can change quickly and deterministically myself, like swapping assets, instead of asking an agent, waiting, and then having to verify the result in-app.

Feels like Replit is doubling down on non-technical operators - people who may not have the time or interest to understand how their apps work even at a high level. Fair game if that’s the ICP that’s working.

But it'd be cool to see more ways to gradually educate users on the key parts of their codebase.

Especially for people who want to get better at building apps, or push their apps beyond a certain level of complexity.

[Joo Tat @heyyyjoo]

@PascalBrokmeier @petergyang Yeah I do this too