For anyone serious about learning to leverage AI

I was just writing on my IG stories and I was thinking deeply about what I would recommend for people if they're serious about adopting AI to subscribe to.

And I came to 2 things.

Some kind of coding tool - chatgpt or claude
Some kind of knowledge work tool - Cleve

This is the stack of the future.

The coding tool subscription allows you leverage the basically infinite capital of Anthropic & OpenAI.

The writing tool gives you a way to embed your knowledge, thoughts and context so you're able to use AI to augment your knowledge work. It's a space for you to think deeper about what you're working on, and output your brain's own tokens such that you can lead and orchestrate an army of agents in the direction of your thinking.

And yeah I'm obviously biased as someone building Cleve, but I'm also betting years of my life into this for a reason and turning down dozens of opportunities too.

Because I think this has to be the future from first principles too.

Notion is the only other company I know that has seen this and is working on this actively. Power users in Cursor, Claude Code and Obsidian are already doing this with awkward workarounds and structures. That's usually a leading indicator that gives an indicator of the products that can change the future.

We've designed Cleve to contain all the features that enables this, and we're not shackled with the branding problem that granola, wispr and cluely have - where they're already pinned at specific use cases. Even Notion has a branding problem in this area where it's more tied to a kanban/productivity tool now.

There's an opportunity for a new space (which I think will contain hundreds of startups) to rapidly emerge. And I think it's a semi-contrarian bet but in a year everyone will realise how important writing / braindumping / documenting human thoughts will become. Your most powerful asset will be your library of documents and thoughts. Your AI models that have access to this will be a natural extension of you and have context of the various opinions and taste you've developed in your brain and allow you to work without the context gap - which is the biggest issue with lots of these tools.

The wave of text/docs apps is coming for knowledge work. And will be the system that really enhances knowledge work more than anything.

The best people in society will be writing and documenting. Then working with an army of agents to work based on this. You can already see the trend of the greatest minds building writing habits. Khailee Ng talks about this, Farza, & many other exceptional people, dating all the way back throughout history. Like most things, this will be amplified with AI.

I want us to be the team that makes this mainstream so we can get this value to society as soon as possible. We've got a lot of work to do but I think we're on the right track.

Some design decisions that I think make sense from 1st principles

People want ownership of their thoughts - no model training, exporting everything is easy, MCP access. High transparency and high trust.
Accessing and depositing ideas in this augmented intelligence should be possible from anywhere - whether it's on your phone, on your laptop, your ipad, on your watch, in your glasses.
Integrations should be widely available so you can connect your knowledge to lots of places.
Provide access to all the models and update as soon as a new one comes in.
Generally have a heavy PLG/Prosumer GTM. Long term games.
Some features need to be designed for viral loops. The winning product will do the above, and add viral loop implementations to win.

Anyways back to focusing on you (the reader) hahaha so what do you need to do to be ahead of the curve (since you're reading this right now)

(my opinions btw as of rn, stc)

Start writing and building your personal knowledge base. It's free with Cleve right now with cloud sync and for the forseeable future. You can do this in Obsidian/Notion if you like paying though.
Create a folder called "Strategy" and put your high level life plans, goals and strategy docs. Very big picture focused. Include a doc called "About me" - so the AI has context.
Start chatting. After writing a piece or braindump, ask Cleve (or any AI chat) its opinions, or expand on your thinking. When working on a problem, use AI to propose ideas or counteropinions on your direction. Ask Cleve based on a week's worth of notes - what ideas do you keep coming back to. Summarise everything you wrote.

It's actually pretty simple and not really that sexy. But that's the reality with most things haha the useful stuff is rarely the most exciting but it's the stuff that gives the most leverage in the long run.

Eventually what you get is an archive of your mind. Which as the AI models (and our harness or structure in guiding the models) get more powerful, they will be able to intelligently read all of it and extend it.

The people who have accummulated this asset will have an unfair advantage generally. But more-so if they've accumulated the practice of having clear thinking and writing over time, and worked on that skill.

Everyone's becoming a CEO of an army of agents in today's world - so if we take that abstraction and look at what the world's leading CEO's do, those are generally the traits to aim for; which revolve around clear thinking, and by extension - writing.

You can start building this today for free with Cleve. Subscribe when you need it and for however long. Use MCP's to Claude Code / Codex / Cowork via the cleve mcp or localsync if you prefer, but we'll be working extremely hard to win in the competition against them to build the best product for these use cases and harness their LLM's intelligence better than the biases that exist in a coding agent.

and I think we have a pretty kickass team to make that happen. join us on the journey maybe? :)

feel free to reach out to me at https://wa.me/60109847954 if you have any opinions, questions or thoughts too, happy to chat!