The full methodology and workflow behind a 22-page, 12-chart research report: Perplexity Computer for data collection and essential research, then Opus 4.8 and Fable 5 for verification and production.
The rule to never publish research from the inside of the machine that produced it is the golden rule of AI generation moving forward. Fantastic breakdown of how to build rigorous quality checks and design discipline into an automated pipeline, John. The 6-section skill architecture is brilliant.
It's intriguing to consider how this approach could transform various fields beyond economics. Which industries do you think could benefit the most from this structured AI research methodology?
Thanks, Melanie! I think there’s huge value in the real estate markets and also in so many B2B sectors with lots of private players who require market research.
The no-hedging voice rule is interesting to me. Once you ban "it seems," a wrong Cisco number reads exactly like a right one, so the whole system leans even harder on the audit step actually catching things. Maybe that's fine since you read every word before publishing anyway.
It’s a great point, Louis. The final pass really is critical but as with so much theses, continuously tweaking the skills and specifics of your directions and prompts is so critical as the models evolve.
"Never publish research from inside the machine that produced it" should be printed on the wall of every AI team. We run research loops every day, and a separate verifier reopens each cited source and fails anything it can't confirm. The brief usually comes back shorter, and we treat that as evidence of the system working.
John, as someone who uses AI for a lot of research, this was incredibly helpful (I do a lot of my trend monitoring in the social sciences this way, including pulling from across academic and consulting houses, then ranking veracity based off source). Creating the system as whole as you've described really creates a powerful machine. Thanks for sharing.
The rule to never publish research from the inside of the machine that produced it is the golden rule of AI generation moving forward. Fantastic breakdown of how to build rigorous quality checks and design discipline into an automated pipeline, John. The 6-section skill architecture is brilliant.
So glad this one resonated with you! Great perspective.
Terrific! I rlly have not explored perplexity
You definitely should!
It's intriguing to consider how this approach could transform various fields beyond economics. Which industries do you think could benefit the most from this structured AI research methodology?
Thanks, Melanie! I think there’s huge value in the real estate markets and also in so many B2B sectors with lots of private players who require market research.
The no-hedging voice rule is interesting to me. Once you ban "it seems," a wrong Cisco number reads exactly like a right one, so the whole system leans even harder on the audit step actually catching things. Maybe that's fine since you read every word before publishing anyway.
It’s a great point, Louis. The final pass really is critical but as with so much theses, continuously tweaking the skills and specifics of your directions and prompts is so critical as the models evolve.
John, this is really helpful...something I am working on building as well.
Thanks for the detailed guide.
Sure thing, Sharyph! Here to help my friend!
AI driven research frameworks are goldmine, thanks for such a detailed write up!
Sure thing, Nirav! 🤓🙏🏻
John, this is brilliant. Do you see this being the workflow for academic researchers as well as economists and marketers?
Appreciate you my friend!
"Never publish research from inside the machine that produced it" should be printed on the wall of every AI team. We run research loops every day, and a separate verifier reopens each cited source and fails anything it can't confirm. The brief usually comes back shorter, and we treat that as evidence of the system working.
John, as someone who uses AI for a lot of research, this was incredibly helpful (I do a lot of my trend monitoring in the social sciences this way, including pulling from across academic and consulting houses, then ranking veracity based off source). Creating the system as whole as you've described really creates a powerful machine. Thanks for sharing.