So I’ve been testing and I’ve come across a very startling observation…
That is to say, that Deepseek is very well set to completely destroy ChatGPT and that’s no joke.
Here’s the thing;
ChatGPT has been on a constant roller coaster of usefulness since it’s inception, and in the sense of wavering from useful, to completely uselessness and back to useful again. And to the point where entire sets of API driven programs are rendered useless.
No such thing with Deepseek
Be it, a baked or distilled model, it’s consistent, which is fundamental to functional
And contrast with, I cannot cound the number of times that Chat LLM’s have undergone recompilation, breaking everything, and rendering projects useless - literally …
NB, right now as I write this, ChatGPT no longer works with most all of my old API Python programs. But that will not be the case with Deepseek, and as I can commit said code, to a particular(existing) model.
BTW, thank-you to the OpenAI team for completely ruining ChatGPT’s capacity to format code text, really nice touch there folks - A+
/rant out