Thanks for the feedback, we'll make sure to reduce the music volume in future videos. Also thank you for pointing out the link! We've updated this now :)
ChatGPT is becoming terrible quicker than CleverBot. It was better when it started but now Errors, including very basic mathematical errors are becoming more frequent. It also more frequently ignores and fails to follow instructions despite also acknowledging which instructions when asked to critique its work and the prompts. It just isn't following the instructions even when you follow its directions for creating a better prompt. In one discussion, i asked it to review a prompt and tell me how many instructions it failed to comply with. 70% was its roughly correct answer. I asked how to better form the prompt and it basically said 'i don't know, I'm just an ai language model' and then gave some tips. Following its recommendations did nothing at all to improve its performance and I should mention, i was asking it to do things it had successfully done in the past and intermittently has succeeded with since then. The last part and following the process in an attempt to gain congruence in the process and instructions only gets it to apologise and explain that it's just a language modelling system. As you try to progress through the contextual enhancements it gets worse and very often repeats the very same mistakes it already identified and explained how to correct, even though you just used the corrected versions.
Thanks for the comment - a lot of folks are reporting a decrease in quality across ChatGPT models (both 3.5 and 4) in the last week. This could definitely be affecting your results. Jump into discord and we’d be happy to help out :)
@@JoshandMarcel Thanks and yes, it's definitely a recent phenomenon (which is part of why it's so frustrating - it's that "you could do this perfectly 2 weeks ago" aggravation).
pls turn down the bg music, otherwise great video!
Thanks for the feedback, we really appreciate it :)
Great topic. Music too loud. Discord invite link to prompt invalid. FYI
Thanks for the feedback, we'll make sure to reduce the music volume in future videos. Also thank you for pointing out the link! We've updated this now :)
@@JoshandMarcel Music is not needed.
I was going to write the same thing, I listen on 1.5 speed and my ears are bleeding. Beside that good vidéo, thank s for your time
thnx for the prompt! the music, however is too loud and u can't b heard properly.
ChatGPT is becoming terrible quicker than CleverBot. It was better when it started but now Errors, including very basic mathematical errors are becoming more frequent.
It also more frequently ignores and fails to follow instructions despite also acknowledging which instructions when asked to critique its work and the prompts. It just isn't following the instructions even when you follow its directions for creating a better prompt.
In one discussion, i asked it to review a prompt and tell me how many instructions it failed to comply with. 70% was its roughly correct answer. I asked how to better form the prompt and it basically said 'i don't know, I'm just an ai language model' and then gave some tips. Following its recommendations did nothing at all to improve its performance and I should mention, i was asking it to do things it had successfully done in the past and intermittently has succeeded with since then.
The last part and following the process in an attempt to gain congruence in the process and instructions only gets it to apologise and explain that it's just a language modelling system. As you try to progress through the contextual enhancements it gets worse and very often repeats the very same mistakes it already identified and explained how to correct, even though you just used the corrected versions.
are you talking about the free version
Thanks for the comment - a lot of folks are reporting a decrease in quality across ChatGPT models (both 3.5 and 4) in the last week. This could definitely be affecting your results.
Jump into discord and we’d be happy to help out :)
@@JoshandMarcel Thanks and yes, it's definitely a recent phenomenon (which is part of why it's so frustrating - it's that "you could do this perfectly 2 weeks ago" aggravation).