Useful system, but the danger is off-loading judgment to a bot. Decision fatigue isn’t fixed by outsourcing thinking — it’s fixed by better frameworks. I’m all for AI structure, as long as the human still owns the call.
love this angle. i've noticed leaders who treat ai as a co-thinker vs just a task tool see way better results. do you find teams resist the shift from doing to orchestrating?
There’s such a wide array of leaders that’s it’s tough to group into one. Older leaders, already successful, often feel are too proud to think anything outside of themselves will help since they need it to get to where they are. Newer leaders or those with a constant learning attitude are more open and usually get the best results, from my experience.
Joel/Jakub - Does the JSON you shared in the article as copy/paste need to be modified in anyway? It looks like you call the OpenAI API using a specific key so I'm curious if someone has to make adjustments to code to ensure they're not inadvertently sending you all of their conversations.
Great article Jakub and Joel and thanks so much for making this process so open and available to us all. Really looking forward to the walkthrough in Part 2! 🙏
The workflow makes sense, but I'm curious about the quality of the AI analysis. Does GPT actually surface non-obvious risks and trade-offs, or does it mostly reflect back what you already told it?
The value here seems to be forcing structured thinking rather than the AI being particularly insightful. Which is fine, that's still useful.
The voice transcription via Telegram is a nice touch for capturing thoughts on the go.
Does the system handle interruptions well if someone starts a new decision mid-conversation?
This is great but AI can't think without having a cognitive layer like CognOS, which you can find here: https://substack.com/@elliotai/note/c-163779830?r=6jttqk
I just love the entire concept of a "thinking partner" 💙
Useful system, but the danger is off-loading judgment to a bot. Decision fatigue isn’t fixed by outsourcing thinking — it’s fixed by better frameworks. I’m all for AI structure, as long as the human still owns the call.
Totally agree! AI should amplify judgment, not replace it.
love this angle. i've noticed leaders who treat ai as a co-thinker vs just a task tool see way better results. do you find teams resist the shift from doing to orchestrating?
There’s such a wide array of leaders that’s it’s tough to group into one. Older leaders, already successful, often feel are too proud to think anything outside of themselves will help since they need it to get to where they are. Newer leaders or those with a constant learning attitude are more open and usually get the best results, from my experience.
Wow I love how accessible you make the decision journal here. Thank you for sharing this! I look forward to the implementation too.
Joel/Jakub - Does the JSON you shared in the article as copy/paste need to be modified in anyway? It looks like you call the OpenAI API using a specific key so I'm curious if someone has to make adjustments to code to ensure they're not inadvertently sending you all of their conversations.
Great article Jakub and Joel and thanks so much for making this process so open and available to us all. Really looking forward to the walkthrough in Part 2! 🙏
Thanks! A much more in depth 2 part piece connected to this drops next week!
The workflow makes sense, but I'm curious about the quality of the AI analysis. Does GPT actually surface non-obvious risks and trade-offs, or does it mostly reflect back what you already told it?
The value here seems to be forcing structured thinking rather than the AI being particularly insightful. Which is fine, that's still useful.
The voice transcription via Telegram is a nice touch for capturing thoughts on the go.
Does the system handle interruptions well if someone starts a new decision mid-conversation?