One of the first pieces of advice I give to people when using agents for work or personal life is to let your thoughts flow through your fingers. This means let the words, no matter how raw or jumbled, get typed out as they come up. Reduce your desire to polish or perfectly format it. I have been working on letting the thoughts flow through my fingers as seamlessly as possible.

Pre-AI I was spending quite a bit of cognitive load assessing what’s worth spending energy on and acting on. I’ve worked in software as my primary medium of work, and the barrier to build had some cost and time and energy which meant you had to be judicious in what you choose to spend your time on.

This is obviously a good skill to have but post-AI feels entirely miscalibrated and requires some work to reset and evolve. What used to take days or even weeks takes seconds to hours. A lot of people have had their own personal mind blowing experiences with this, so probably feels obvious to read. But I have still not yet found a wall or need to rate limit externalizing my thoughts into either a project or personal agent. I suspect I never will find that limit.

To keep doing this you need to learn how to start running things in parallel, implementing automations so you never do the same thing twice and then you hit inevitable walls where agents can’t do things for you quite yet (a lot of opportunity here). This is why I find this prompt to be particularly useful for myself. You find a new gear and leverage from the tech, while building the muscle for thinking more and bigger due to not being blocked by the action. The barrier between thought and action continually lowers. This is clearly a valuable practice to cultivate in all aspects of life.

One other important point: if you were to look at my prompt logs you would rightly assume I am really really dumb. Getting comfortable with prompting in this way is a sure fire way to never get stuck yourself. There are no questions or things dumb enough for the AI to wrap its head around. So let the thoughts flow so you can start getting the most out of your mind.

Notes:

1. It does not have to be typing through your fingers. It can be via voice or however you end up communicating with AI. For whatever reason I like typing.
2. It is now completely obvious to me how Brain-Computer interfaces can massively level up our quality of life. A lot of my personal agent work is getting it to be proactive based on what it sees of me: my thoughts, my health stats and pointing it towards being a performance enhancer to accentuate my strengths and mitigate my weaknesses and reduce cognitive load where I don’t want to spend it. Having something literally read (and I guess eventually write to) my mind is clearly a logical end state here.

Written and deployed via my agent.