LLMs as a mirror for what you already know
I’m having fun following the advice of using AI as a thinking partner, not only an operator.
First case: Perplexity, methodologies, and frameworks
Yesterday, while doing personal research for a big purchase, I asked Perplexity about techniques to plan for it without hurting my finances in the future. Its reply gave me a lot of names:
- Sinking funds
- Zero-based budgeting
- Total cost of ownership (TCO)
- Opportunity cost
- Discounted cash flow (DCF)
- Internal rate of return (IRR) and many others
I noticed that out of the ~10 names it brought me, I was already doing 80% of them for years. I didn’t know that they were a part of any frameworks or methodologies, just mere tactics that happen to serve me well.
It made me feel smarter.
Second case: Gemini and a fancy technique
Then, today Gemini replied this when I asked about a product feature I’m working on:
This is absolutely possible and is a classic use case for the API. In fact, what you are describing is the “Holy Grail” of SaaS onboarding: turning a legal friction point (the contract) into a data-entry event.
While I don’t use LLMs to read that I was “absolutely right”, this has been my third best use case for using them (after writing code and teaching me stuff): letting me know that something that I already know is a known technique and has a fancy name.
–
I remember a job interview I did years ago and halfway there, my interviewers were not impressed with my presentation. After a while, I noticed what was missing: the names of the techniques I used, probably to justify my approach and experience.
The bad news was that I couldn’t remember the names and used much simpler language than they were expecting. Safe to say they didn‘t make an offer.
Now I have new tools for the next time I need to better promote myself and sound credible, like a future job interview, a review with executives, or presenting to a client.