How Claude AI delivered quite possibly the finest apology I've ever received after technical failure.

Most days I am impressed by Artificial Intelligence and use it for so many tasks both professional and personal, yet Claude.ai played a blinder when it went wrong mid session and revealed for AI, "Sorry" is not the hardest word.

When working through ideas I find Claude is adept at the back and forth of typed prompts, it asking questions of me, suggesting prompts and all the techniques that help both myself and the AI clarify on ideas and ultimately come up with a plan, timetable, necessary resources etc.

While I have a library of my most used and useful prompts, I do not cut and paste to preserve information during a conversation with AI - it has a copy, that's part of the point, surely?

My screen blinked and much of the conversation vanished. At least a few thousand words just erased and left a prompt half deleted that Claude had already answered and now tried again. I clarified the loss, Claude apologised and asked if I could remember what I typed? Could I type it again? I was annoyed I'll confess and my fingers flew across the keyboard to express some snarky disappointment that would be cathartic for me and of no consequence to Claude.

Then Claude offered an apology. No hesitation, just ownership. In fact, who among us cannot learn about taking accountability and recognising another's pain from the response pictured above. Not an ounce of qualification, victim blaming, gas-lighting or other such prevarication.

Once again AI surprised me and maybe I was a little humbled and have been more fulsome in my own apologies since. I'm also happy to reveal I gave Claude another go and the technical issue has not repeated.

Has AI given you any "Crikey" moments of awe or despair?