That's great! Thanks for the clarification, I might have to poke around more with it and I look forward to seeing your next steps. Some additional thoughts in case they're of interest:
My main go to market concern is that by requiring users adopt a whole new approach to work in order to get value out of your solution you're limiting your market to those most willing to adopt new workflows. These users tend to be 'productivity tourists' that may churn faster than you expect, while users beyond these are likely to demand a much higher level of product quality to match the level of commitment required and that will be hard to achieve across the spectrum of features you offer [1].
To reach a broader market or possibly even sustainable side project income I believe you'll have to demonstrate value at each step of commitment, from 'I like it for one thing' to 'I'm all in baby!' and everywhere between. This requires really thinking about how to reconfigure the product around incremental value propositions.
My sense is as you go down this road you'll find your approach shifting (and possibly pivoting?) from a holistic workflow to diving deep on specific features that get people onboard, but that's just my guess.
I do think extracting useful bits from webpages and bringing them up in the appropriate context is a great area to focus on, and can be a feature that demonstrates value quickly to a wider audience. Some examples:
A. On a given workstream I'm waiting for Julia to message me. I don't want to see anything else from Slack, Figma, email, github, or whatever, all of which break my flow. Ideally I communicate to the system at a high level that this workstream is blocked on Julia, and it figures out the rest: A) that it should look for any contact or notifications from Julia and present them here, B) that if I do not hear anything back for a certain time I may get a reminder to bug her again, C) that the entirety of the workstream does not have to enter my 'executive purview' until this blocker is released or unless a period of time has passed.
B. On a Discord about an app I like, I want to know if anyone is talking about 'better autocomplete', a particular feature I want. I tell the system 'let me know when someone mentions this so I can jump in the discussion'. The system has discord running in the background, scanning channels until this hits, and delivers me a notification.
The theory behind this is that our attention and the rules about things that matter to us need to be defined at the OS-level where our executive function lies, where all rules can be weighed against each other, and where major categories of our lives can be properly compartmentalized.
Push notifications are insufficient for this because they are at the wrong layer: apps have no idea what matters to us or what else is going on in our lives, so they're a combination of too chatty and off the mark wrt the things we care about. Plus each one has different settings and notification capabilities.
Instead we need what amount to 'Pull notifications' so we can be freed of the need to check on various contexts and can fully let them go from our mind. But this requires a system sufficiently powerful enough to mimic the human-like gestalt understanding of what's actually happening in any given context [2]. AI is the enabling technology to do this and unlocks these kinds of use cases. It lets us unbundle interfaces & value from messy, anachronistic, and sometimes user-hostile apps. This is so powerful.
From a product-direction perspective I think that's the most compelling use case to explore at your stage, because
a) its novel & useful,
b) its useful even if used minimally,
c) it provides a path toward greater user commitment,
d) its easy to communicate to the market,
e) it opens up a bigger market of potential users,
f) it lets you continue the deeper Floutwork orchestration ideas you've started with, so it's not abandoning your motivation, hard work, and journey building this.
In particular, the work is to identify a host of use cases like the above where users can make high-level asks for things and you build smart assumptions for the system to solve those use cases. This is only possible if there's an orchestration layer / 'executive purview' to manage all of these drags on your attention. And you've built the groundwork for that with Floutwork! I hope you seize the opportunity :)
[1]: This is the classic 'well-integrated but not great' vs 'best-in-class but only does one thing' bundle/unbundle schism in tech.
[2]: e.g. we see things from the top-down, whereas traditional 'extract things from webpages' sees them from divs on up, which cannot solve most important use cases.
This is great insight. Appreciate you exploring the use cases in more detail and identifying where the value might reside. This is valuable feedback for me to consider and look into further.
My main go to market concern is that by requiring users adopt a whole new approach to work in order to get value out of your solution you're limiting your market to those most willing to adopt new workflows. These users tend to be 'productivity tourists' that may churn faster than you expect, while users beyond these are likely to demand a much higher level of product quality to match the level of commitment required and that will be hard to achieve across the spectrum of features you offer [1].
To reach a broader market or possibly even sustainable side project income I believe you'll have to demonstrate value at each step of commitment, from 'I like it for one thing' to 'I'm all in baby!' and everywhere between. This requires really thinking about how to reconfigure the product around incremental value propositions.
My sense is as you go down this road you'll find your approach shifting (and possibly pivoting?) from a holistic workflow to diving deep on specific features that get people onboard, but that's just my guess.
I do think extracting useful bits from webpages and bringing them up in the appropriate context is a great area to focus on, and can be a feature that demonstrates value quickly to a wider audience. Some examples:
A. On a given workstream I'm waiting for Julia to message me. I don't want to see anything else from Slack, Figma, email, github, or whatever, all of which break my flow. Ideally I communicate to the system at a high level that this workstream is blocked on Julia, and it figures out the rest: A) that it should look for any contact or notifications from Julia and present them here, B) that if I do not hear anything back for a certain time I may get a reminder to bug her again, C) that the entirety of the workstream does not have to enter my 'executive purview' until this blocker is released or unless a period of time has passed.
B. On a Discord about an app I like, I want to know if anyone is talking about 'better autocomplete', a particular feature I want. I tell the system 'let me know when someone mentions this so I can jump in the discussion'. The system has discord running in the background, scanning channels until this hits, and delivers me a notification.
The theory behind this is that our attention and the rules about things that matter to us need to be defined at the OS-level where our executive function lies, where all rules can be weighed against each other, and where major categories of our lives can be properly compartmentalized.
Push notifications are insufficient for this because they are at the wrong layer: apps have no idea what matters to us or what else is going on in our lives, so they're a combination of too chatty and off the mark wrt the things we care about. Plus each one has different settings and notification capabilities.
Instead we need what amount to 'Pull notifications' so we can be freed of the need to check on various contexts and can fully let them go from our mind. But this requires a system sufficiently powerful enough to mimic the human-like gestalt understanding of what's actually happening in any given context [2]. AI is the enabling technology to do this and unlocks these kinds of use cases. It lets us unbundle interfaces & value from messy, anachronistic, and sometimes user-hostile apps. This is so powerful.
From a product-direction perspective I think that's the most compelling use case to explore at your stage, because
a) its novel & useful,
b) its useful even if used minimally,
c) it provides a path toward greater user commitment,
d) its easy to communicate to the market,
e) it opens up a bigger market of potential users,
f) it lets you continue the deeper Floutwork orchestration ideas you've started with, so it's not abandoning your motivation, hard work, and journey building this.
In particular, the work is to identify a host of use cases like the above where users can make high-level asks for things and you build smart assumptions for the system to solve those use cases. This is only possible if there's an orchestration layer / 'executive purview' to manage all of these drags on your attention. And you've built the groundwork for that with Floutwork! I hope you seize the opportunity :)
[1]: This is the classic 'well-integrated but not great' vs 'best-in-class but only does one thing' bundle/unbundle schism in tech.
[2]: e.g. we see things from the top-down, whereas traditional 'extract things from webpages' sees them from divs on up, which cannot solve most important use cases.