That’s an interesting perspective. However, the internal persona the model adopts is a different area from the one I was exploring. My focus was whether pleasantries affect output quality and how to elevate prompts to be more efficient, all while preserving tokens.
That said, what happens inside is worth exploring separately. Do you have an article to suggest?
I believe the best way is to call them out on it. If the number of these failures keeps rising like that, there’s a risk of it becoming a regular occurrence. Whenever you see it, point it out by commenting on the post itself or share it with your thoughts, or write in your blog like I did. That will make them double and maybe triple check next time.
Hey, Moneer, great to hear your views on this
To be thorough, I did point out the mistake in the Africa image to the page on LinkedIn that published it, and asked them to tell me which AI they’d used and the prompt. They didn’t respond.
My CELTA tutor told me once that many trainees would submit written assignments with grammatical and structural errors, and these are always underlined in the word doc in red, no less.
It would be great to really know why someone skips checking AI outputs, given that it’s prone to hallucinations. They might be the same ones who submitted to my tutor :)
Until then, we just need to be watchful and call out the mistakes. Maybe someone will pay more attention because of it.
First of all, thank you for the post, great effort on your part that doesn’t go unnoticed.
That said, when I hear talk of talent, it’s usually talent located in the U.S. and EU (specifically the UK) where most fellowships and jobs are offered—the global majority is not included. The very few that are open to global applicants are for technical individuals.
What about researchers who have something to contribute but face so many closed door?
This LinkedIn post is of a recent experience I had during a webinar, which might bring what I’m trying to say home.
Thank you so much for your comment and for participating in the survey, appreciated :)
Thank you so much for posting this. I too found the link broken
Thank you so much
This is a valid point, but in a consciousness debate. The topic here is different. The responses are direct answers to a direct question—the quality of the output if no pleasantries were used.
A point they all made was ‘not creating noise’ - emotionally-charged language gets in the way. This would make it difficult for AI to do its actual job of responding to prompts.
For my take on the consciousness debate, see here