Quote Originally Posted by Niya View Post
Here's the thing though, I've rigorously tested the programming capabilities of 3 different iterations of GPT and each successive iteration massively out-performed the capabilities of the previous iteration. From that I could extrapolate how fast AI is evolving and where it's likely going to be in the next year. It may be true today that you will need to be a programmer to use it effectively as you will have to correct it's mistakes but I doubt that will be the case next year. I would not be surprised next year if it stops making mistakes altogether. Eventually, it will get good enough to where even non-programmers would be able to get solid outputs from it.

I am telling you, AI is evolving rapidly and I do mean rapidly. I have never in my life seen anything like this. AI is making leaps and bounds in timescales that could be measured in months, not years or decades as is typical with technology.
You may be right, we'll see.

But a lot of things make large leaps in the the beginning. Which makes sense because there are large areas for improvement. Fixing the last 10% can take longer than creating the first 90%.