Thanks for writing this, Charles — really enjoyed your reflections!
Quick question: is the paper you reference in the appendix (“offering structural definitions about how OpenAI classifies user behavior…”) the “How People Use ChatGPT” paper by Chatterji, Cunningham, et al.?
When I learned to code in Ruby on Rails about 15 years ago, I found the exercise immensely valuable because it forced me to think (1) systematically and (2) completely about what I was trying to accomplish. I hope that AI leads many people to have the experience you describe here, because it is such a new and fascinating way to look at the world.
Love your latest post - it really helped me understand a bit more the nuances around how people are using and evolving these tools they're building. You've really jumped in!
Valuable insights! Especially the part on making implicit choices explicit.
One part was covered- you were automatically assessing as before, regardless of the new data that you thought you incorporated in your thought process.
En there, this exercise helped.
On the other part, I am curious to hear, what happens with intuition?
What if what we know is not only what we think we know?
In that case, we will never be able to create a perfect set of instructions nor fully grasp what makes us great in our job, because there are hidden fine processes that happen, that have been scientifically proven to be very much valuable and come with yrs of experience?
That doesn't mean one should stop with trying to figure out their investment decision making process and all the logic and systems they use.
I am just saying, let's leave room for errors even in this process, since there is inherent limitation to how far we can go.
Thanks for writing this, Charles — really enjoyed your reflections!
Quick question: is the paper you reference in the appendix (“offering structural definitions about how OpenAI classifies user behavior…”) the “How People Use ChatGPT” paper by Chatterji, Cunningham, et al.?
Yes, it's that one!
When I learned to code in Ruby on Rails about 15 years ago, I found the exercise immensely valuable because it forced me to think (1) systematically and (2) completely about what I was trying to accomplish. I hope that AI leads many people to have the experience you describe here, because it is such a new and fascinating way to look at the world.
Love your latest post - it really helped me understand a bit more the nuances around how people are using and evolving these tools they're building. You've really jumped in!
Valuable insights! Especially the part on making implicit choices explicit.
One part was covered- you were automatically assessing as before, regardless of the new data that you thought you incorporated in your thought process.
En there, this exercise helped.
On the other part, I am curious to hear, what happens with intuition?
What if what we know is not only what we think we know?
In that case, we will never be able to create a perfect set of instructions nor fully grasp what makes us great in our job, because there are hidden fine processes that happen, that have been scientifically proven to be very much valuable and come with yrs of experience?
That doesn't mean one should stop with trying to figure out their investment decision making process and all the logic and systems they use.
I am just saying, let's leave room for errors even in this process, since there is inherent limitation to how far we can go.