Summary
Take the tasks that we normally give to MBA Interns in a startup (“Research these companies and produce a stack-ranking of which we should talk to first”, etc), and try via LLM instead
Usefulness: Practical/Productive
Techniques
- Very open-ended; compared to a smart MBA, LLMs are both amazingly good and hilariously bad; requires a lot of “try something, if it sucks – discard it entirely”
Discoveries
- NB: this (failed) experiment later inspired the (successful) experiment – https://aipioneerspath.com/2024/05/14/experiment-get-a-world-expert-generate-nuanced-high-quality-sales-material-for-new-products/
- Flip the table: ask for a hallucination of what the world looks like to the target stakeholder, then mine those for insights to help me re-think / re-vision / re-target my own (outside of LLM) activities – pretty good results so far