Allegedly, the advent of AI means that there will no longer be such a thing as a typo. Yeeha. So, no more “genital reminders”, “speak shorty”, “very busty” or wishing someone has a “lonely weekend” in work emails.

AI spell-checkers are brilliant. Where a ‘normal’ spell checker won’t pick on pubic where it should be public, for example, an AI one will. And, we can all benefit from a vigilant personal copy editor, someone to highlight any glaring errors before they become mortifying emails to colleagues or worse, an external email or worse still, an entire email list.

That said, there’s something endearingly human and chuffing hilarious about a good typo. It’s never not funny reading about corporate, or other, typo gaffes.

So, will AI stop these? I actually think not.

Whilst I do think there might be fewer typos in weighty tomes like annual reports/investor docs, I reckon there will always be typo slip-ups.

It has always baffled me that road markings are ever wrong, but I suspect it will be ever thus, AI or no AI. Would AI have picked up a NO ENRY? I say no. There’s also been a recent “CEEP KLEAR”. There is no shortage of similar worldwide, and I think there always will be. Road markings aren’t typed into a document that gets spell-checked. They’re painted by actual humans who have to manually spray-paint letters onto tarmac. Unless AI comes at the end of a paintbrush, I suspect these will always happen.

Bristol City Council recently spent nearly £200,000 sending out 239,143 letters to explain a typo on council tax bills. They’d transposed two percentage figures – the Avon Fire Authority and Police and Crime Commissioner precepts were listed the wrong way round. The bills were correct; just the percentages were swapped. But by law, they had to write to everyone to explain.  £200,000. For a typo. That AI didn’t catch. Because the error wasn’t in the typing – it was in the data entry, the proofreading process, the chain of approvals. Human error, layered upon human error, in systems too complex for a simple spell-checker to fix. So yes, the potential for more £200k cock ups is still around.

Another one of my ‘cannot believe it got printed’ bugbears is product typos. Multiple people see these designs at various stages of the design to print to product in store process. And yet they slip through. AI might check the copy, but it won’t catch the disconnect between what someone typed and what they meant, or spot when a product description doesn’t match reality. That said, any old spell checker should have picked up this one – starwberry is NOT a word. 

So no, I don’t think AI will mean the end of typos. I think it might mean a different kind of wordy faux pas – an over-reliance on AI leading to contextual cock ups or just verbal diarrhoea that’s meaningless waffle. AI will save us (potentially) from the kind of error a local school made a few years ago, when sending a wet-weather text to parents who were asked to ensure kids brought their willies to school (wellies, obviously). I can’t find a link now, but I assure you it did happen. AI adoption is an opportunity to make our communications cleaner, clearer, and more professional. But typos will always exist, plus, I suspect we’ll find new ways to mess things up spectacularly.