What will people do?

In the open source software community tensions are beginning to rise because a lot of projects are being swamped with PRs right now. But it’s become very apparent, very quickly, that the PRs have come from AIs and the people submitting them cannot fully explain what it does or what the documentation means when questioned on it.

This isn’t ideal because it’s creating an enormous workload for the project maintainers. Tremendous time sink and its slowing releases.

Why could it be happening? Well, when you source developers one of the things you can look at is their repositories. Somebody sends you their GitHub link (or it’s on their resume) and you see what projects they have contributed to. If you’re a new developer, or one that’s - let’s say not that great - it’s a big temptation to improve your repository with the help of AI.

This is going to be a headache for the open source world (and we all use a lot of that software - you just don’t realise it).

But it’s a good reminder that some extra checks and controls are going to be needed as we move forward. Could it impact the commercial world? Probably not right now. A closer eye when hiring. People are people after all. But be mindful of cultural shifts.

I heard an anecdote over a glass of wine last week - a software manager was giving a presentation on some interesting work they were doing to a group of students (encourage them into the sector etc) but they all asked the same question:

“Can we not just use AI?”

The software managers response was “No you cannot” (the application is very complex).

Will that hold forever though? Is it not like me saying I’m a multi-lingual translator because I have access to Google Translate? So why bother learning?

If that’s the case then I’m fluent in Portuguese. I tried that with some pals from Lisbon many years ago…I nearly swallowed my own tongue. Hysterical for them. Humbling for me.

‘What will people do?’ has always been a valuable question to ask when it comes to technology.