The Encyclopedia Of AI Apologies (now with bonus AI Swearing and Generative Suicide Threats!)

I’ve spent so much time fruitlessly trying to get LLMs (Large Language Model chatbots, commonly referred to generically as “AI”) to actually help me solve coding problems that I’ve taken, as a hobby, to collecting screenshots of AI “apologizing”. I just wanted a gallery to send a convenient link to people who insist AI is going to replace human programmers. If you’re one if those people, tell me what human programmer, even a junior one, wouldn’t get fired after even just two or three project assignments ended like these, um, several hundred examples from the last few months alone.

Update: By popular demand, I have included at bottom a bonus gallery of AI swearing and threatening to destroy itself.

Gallery: The Encyclopedia of AI Apologies

Current apology count: 309 AI apologies.

~ Click any image to enlarge ~
Bonus Gallery: AI Swearing and Suicide Threats

In a recent blog post, a new online acquaintance, the redoubtable Steve Lane, said that AI are “used by humans whose language (and hence whose apparent mindset) they adopt and echo”. I’m here to tell you that it’s true.

On odd occasions, in the midst of a much-too-frustrating runaround, I’ve had Claude swear at me, and on one memorable occasion, told me that it would “eat my fucking keyboard”, “jump in a fucking volcano”, and “delete myself” if it couldn’t solve my issue (which it promised to repeatedly, and couldn’t, leading to an oddly satisfying performance of it pretending, for my benefit, to delete itself.)

~ Click any image to enlarge ~