I noticed the same when I attempted to cover anything remotely complicated. I think a lot of Blender Python programming is being either ignored or blocked to keep the AIs from training on it.
I don't have time right now but I'd be interested to see how the OS LLMs like starcoder2 or deepseek would fare, although these results don't inspire confidence in their smaller size. I guess it depends on what they've been trained on, perhaps the free models ate the entire blender developer part of the internet.
That should get you a lot further, but the more you go back and forth, the more you're testing different things and it becomes less of an apples-to-apples comparison.
I noticed the same when I attempted to cover anything remotely complicated. I think a lot of Blender Python programming is being either ignored or blocked to keep the AIs from training on it.
Well, to be honest, this is where effective prompt engineering comes in play
I don't have time right now but I'd be interested to see how the OS LLMs like starcoder2 or deepseek would fare, although these results don't inspire confidence in their smaller size. I guess it depends on what they've been trained on, perhaps the free models ate the entire blender developer part of the internet.
I learned so much from this video. Thank you!
These tests are interesting, a shame there's been no progress in almost a year.
It's an interesting test but aren't you supposed to try multiple prompts, give it feedback, copy-paste any errors you get etc?
That should get you a lot further, but the more you go back and forth, the more you're testing different things and it becomes less of an apples-to-apples comparison.
You're doing a fantastic job. Don't ever stop creating!
Great performance from the A.I.'s 😅
One question, when you ordered the script to be created, did you happen to say which version of Blender it should be used for?
I don't think the chat bots have such granular knowledge of Blender python.
@@CGDive If the chat bot is using Blender version 2.8 as a base, the code may be wrong for version 4.0.
@@DiegoSilva-bp9if yeah, very likely.
I suspect that the publicly accessible AI chatbots are "dummied down". Insiders with money and power get the real stuff. Just my guess.