Here's the most 2018 sentence you'll read today: Google's email service, Gmail, has removed gender pronouns from its Smart Compose email settings after the artificial intelligence behind it developed sexist tendencies.
OK, so that sounds a bit like the initial plot of a really boring science fiction film, but it's actually true.
It probably wouldn't be as good as the Scrubs episode written by an AI bot, anyway.
Google noticed that the Smart Compose feature, which is designed to make life easier for anyone emailing by trying to guess in advance what they'd like to say, was accidentally using gender pronouns that could have caused offence.
With that in mind, the company has decided to stop doing it altogether. You can still use the feature, but instead of things like 'she' or 'him' it will now offer you things like 'it', 'you', or 'they'.
This change comes after it was discovered that the AI was wrongly assuming gender in a way that could - if it were actually a living being - be considered sexist. Weirdly, there is a growing concern that some AIs are becoming predisposed to sexist tendencies.
Google is removing gender pronouns from Gmail's AI. Credit: PA
In reality, though, it's not about 'political correctness gone mad' - it is about correcting potential mistakes. Remember when you called your teacher 'mum' at school? Like that, but not funny.
The problem was brought to light by a research scientist at the search giant who noticed that when he sent an email with the text "I am meeting with an investor next week", Smart Compose suggested that he should ask "Do you want to meet him?" as a follow-up question.
Obviously, we're all used to the occasional auto-correct or text gaffe, but gender is a hot topic now and it could be damaging to Google's business if it is consistently getting it wrong.
Gmail product manager Paul Lambert told The Independent: "Not all 'screw ups' are equal."
He added that gender is a "big, big thing" to mess up. Especially when you consider that Gmail has 1.5 billion users worldwide, and Smart Compose accounts for 11% of the messages sent. Even at a small percentage of that being wrong, that's a whole lot of mistakes.
There are enough mistakes being made in the world of AI as it stands. Check this out...
However, if Google shows that it can cut out those mistakes then it will have demonstrated that it is leading the way with regards to AI development.
That being said, it did eventually turn out that in order to limit the number of mistakes it was making, the only way around it was to just limit the use of them altogether.
Lambert said that the changes actually only affect around 1% of emails sent using Smart Compose. If that means that nobody is getting offended, and suits Google's aims, why the hell not?
Featured Image Credit: PA