someone needs to tell these "AI" kids to shut the fuck up
You are not a "prompt engineer" bro ur just becoming socially aware
Listen, I’m sorry. Controversial perhaps but it needs to be done. I will not allow you fucking nerds to keep using words incorrectly. This is why everyone hates STEM it’s cuz you people do not know how to look up things in dictionaries!!!! EVEN THO THAT IS LITERALLY ALL PROGRAMMING IS IS LOOKING UP STUFF IN THE DOCUMENTATION YALL KEEP FORGETTING TO LOOK AT THE DOCUMENTATION OF THE ENGLISH FUCKING LANGUAGE!!!!!
like look at this fucking guy right here. He has this list of “prompt secrets” which like are literally just how to have good communication (MAYBE IF YOU PUT THIS MUCH THOUGHT INTO TALKING TO OTHER POEPLE YOU WOULDN’T HAVE TO SIT AROUND TYPING AT A ROBOT ALL THE TIME) but like, for every single “secret” all he does is define some abstract concept and then asks ChatGPT to “break down” an example prompt into elements revolving around the concept and he DOESN’T EVEN PROVIDE A SINGLE COUTNER EXAMPLE like he just goes “as you can see, ChatGPT was able to identify the ‘context’ of this prompt, so that means im right.” Which like, yes, if ChatGPT is a “language model” trained on “text” rendered in “human language”, and meaning in “human language” is literally constructed thru CONTEXT then yes, it is literally 100 percent natural that the fucking robot would have more relevant output when provided with additional “CONTEXT CLUES.” But like, he doesn’t even do the basic fucking STEM thing of trying to replicate your results like even ONE SINGLE TIME.
ABD THEN, FOR TIP 4, HE STRAIGHT UP DOESN’T PROVIDE A SINGLE EXAMPLE!!!! he talks about this thing which he calls “iteration” and like, he’s not even talking about making prompts redundant so that ChatGPT has multiple opportunities to pick up on your intended instructions, he’s fucking talking about REVISING THE PROMPTS WHEN THEY DON’T WORK HOW YOU WANT THEM TO@! Seriously go to 6:53 in that video and LOOK at what he has written on the screen.
“Iteration is a crucial part of the prompt design process” actual corporate jargon
“Prompt design is an iterative process and may take multiple attempts to create an effective prompt” literally all things are like this???? you literally do not get anything in life right on the first try!!!!
“Each iteration should be guided by the results of previous trials, aiming to address any limitations encountered” duh??????????
And then tip number 5 is just “how to use these 4 tips to create a prompt” its like he literally calls the video “5 secrets” like actually chill urself my guy. And they (dumbass AI evangelists) are all like this!!!! and of course the comments are FILLED with people going “oh my god this is so neat” and its like why the fuck can’t I convinced a bunch of dumbasses to think that EYE am special, hmm?????????
Don’t get me wrong, it’s not like there is nothing to the idea of “prompt engineering.” ChatGPT is a tool, and like all tools, there is an art to utilizing it effectively. But like, I wouldn’t consider ANY of the things this fucking guy talked about in his video to even approach the idea of “manipulating” or “extracting” information from ChatGPT. All you boneheads are doing is just learning how language works!!!! and, especially when videos like this motherfucker where like, there is LITERALLY A SALARY FIGURE IN THE VIDEO and its like BRO how am I supposed to take any of this shit seriously?????
idk maybe at some point I will do a deeper dive into this shit because like, idk man this shit is so fucking goofy for real. it’s like i asked ChatGPT to write me a story about how humans don’t understand what ChatGPT is!!!