✨ Nora Tindall  🪐

Well, an AI understands what "usually" happens when something happens That's basically how we work too. We learn from past experiences and learn to react accordingly.

You're not being precise when you say "an AI copy/paste something". It does not. It reads millions of examples and "learn" how to do something.

It isn't copying a given chunk of the code from somewhere. It is mixing everything that it read and creating a given code.

So, when you see a code chunk generated from an AI algorithm, it is basically impossible to define from where it came from. Like it is impossible to define from where you came up with the idea to code what you just coded this morning. You learned from past experiences.

You are literally comparing a human's ability to learn from past experience with the kind of AI technology used in GitHub Copilot. I really didn't know we already lived in a science-fiction novel! It's going to take decades of new improvements before it comes any close

Well, read a little about GPT-3 I'm not comparing humans to machines, I'm just saying that both learning processes can be compared, precisely because machine learning was inspired by the human learning process

@awakecoding I'm going to take a different approach trying to explain to you how I see this. Consider the following example (as I've said, this is ALREADY happening): we get the GPT-3 transformer and train a neural network to write movie scripts. To specialize our (+)

To specialize the GPT-3 network in our domain we're going to use thousands of movie scripts. We're using movies from Disney, Terminator, Start Wars. A lot of licensed movies (this is already happening, as I've said) After training our network, it generates a movie script (+)