I often feel misunderstood, so what if I A/B tested this...
I write a blog post about something which is very niche or personal and thus bound to get misinterpreted. I try, to the best of my ability, to make it as clear as possible and post it here.
Then, I take that exact post as posted, put it into an LLM and ask it to rewrite it and keep asking it to rewrite it until I feel it expressed what I intended to express. Maybe a couple of LLMs and duke them out.
Finally, I post it on here.
I’m not sure how I would test which one is more successful? Enthusiasm (positive or negative emotions) of replies? Positively scoring the post with more replies I feel are most “on the mark” and by extension, negatively scoring the post which has more irrelevant or tangential replies?
I would have some disclaimer that, yes, they are the same exact post even if they are written differently—or would that foul up the experiment? Not sure if I will identify which one is the LLM one and which one is my original.
{How meta is this—I asked Claude to rewrite the above}
When I speak, people misunderstand—they appear to construct an entirely different narrative, responding passionately to a statement I never made.
My goal is simple: find a way to transmit my exact meaning with precision. Can another voice render my thoughts in a way that ensures the core intention is correctly understood by the audience on the blogging platform?
The critical question becomes: How would I determine which version—my original or the translated piece—more effectively communicates my intended meaning? Would I measure success through the enthusiasm of responses, score the replies based on their relevance, or track the proportion of comments that genuinely reflect the original message? Perhaps I’d assign positive weight to responses that demonstrate true understanding and negative weight to those that drift into tangential or misaligned interpretations.
The meta-question becomes: Can a different articulation—whether by AI or human—create a pathway to being truly heard, or are misinterpretations inevitable regardless of how carefully the message is constructed?
{Then I asked it to do it like a pithy Hemmingway… because… why not...?}
People twist my words, inventing stories I never told.
I want precision. Can another voice make my meaning clear to readers?
How to measure success? Count relevant replies. Score understanding. Punish misinterpretation.
Can a different voice break through, or is misunderstanding inevitable?
I often feel misunderstood, so what if I A/B tested this...
I write a blog post about something which is very niche or personal and thus bound to get misinterpreted. I try, to the best of my ability, to make it as clear as possible and post it here.
Then, I take that exact post as posted, put it into an LLM and ask it to rewrite it and keep asking it to rewrite it until I feel it expressed what I intended to express. Maybe a couple of LLMs and duke them out.
Finally, I post it on here.
I’m not sure how I would test which one is more successful? Enthusiasm (positive or negative emotions) of replies? Positively scoring the post with more replies I feel are most “on the mark” and by extension, negatively scoring the post which has more irrelevant or tangential replies?
I would have some disclaimer that, yes, they are the same exact post even if they are written differently—or would that foul up the experiment? Not sure if I will identify which one is the LLM one and which one is my original.
{How meta is this—I asked Claude to rewrite the above}
{Then I asked it to do it like a pithy Hemmingway… because… why not...?}