It probably does. But an amazing number of commenters think they are prompting the copy & pasting, and hoping for the best.

Yep, a lot of headline readers here.

It's just a very advanced autocomplete, completely integrated into the internal codebase and IDE. You can read this on the research blog (maybe if everyone just read the blog).

e.g.

I start typing `var notificationManager`

It would suggest `= (Notification Manager) context.getSystemService(NOTIFICATION_MANAGER);`

If you've done Android then you know how much boilerplate there is to suggest.

I press Ctrl+Enter or something to accept the suggestion.

Voila, more than 50% of that code was written by AI.

> blindly committing AI code

Even before AI, no one blindly accepts autocomplete.

A lot of headline-readers seem to imagine some sort of semi-autonomous or prompt based code generation that writes whole blocks of code to then be blindly accepted by engineers.

That makes a while since I’ve done Android, but I’m sure that this variable should be a property and be set as part of the lifecycle. And while Android (and any big project) is full of boilerplate, each line is subtly different or it would have already been abstracted in some base class. And even then, the code completion is already so good in Android Studio that you would have to be a complete junior (in this case, you wouldn’t know that the AI suggestion is good) to complain that writing code is slow. Most time spent is designing code, fixing subtle bugs, and refactoring to clean up the code.

> The other 75% is the stuff you actually have to think about.

I’m pretty sure the actual ratio is much lower than that. In other words, LLMs are currently not good enough to remove the majority of chores, even with the state of the art model trained on highly curated dataset.

Have you used CoPilot with vs code? It's not perfect all the time but its autocomplete is right a significant amount of time.