Look, I didn’t want to write this piece. I was perfectly content nursing my third bourbon of the morning, contemplating the metaphysical implications of my latest hangover. But then this gem landed in my inbox, and well… here we are.
OpenAI, those wonderful folks who brought us ChatGPT and a whole new way to plagiarize college essays, just pulled what might be the most expensive “dog ate my homework” excuse in recent memory. They managed to delete crucial evidence in their ongoing legal battle with the New York Times and Daily News. And not just any evidence - we’re talking about the very data that might prove whether they’ve been stealing content like a drunk guy at an all-you-can-eat buffet.
Let me set the scene while I pour another drink.
These news organizations are suing OpenAI for allegedly scraping their content without permission. Fair enough. OpenAI, being the generous souls they are, provided two virtual machines for the plaintiffs’ lawyers to search through their training data. Simple enough task, right? About as simple as me remembering where I left my keys last night.
Here’s where it gets good: On November 14th, some engineer at OpenAI hit the wrong button or spilled coffee on the wrong keyboard, and poof! All the publishers’ search data on one of those machines vanished faster than my last paycheck at the liquor store.
Now, I’ve deleted my share of important files. Hell, I once deleted an entire article about data security while drunk-typing at 3 AM. But I’m not sitting on a $90 billion valuation and pretending to be the future of humanity.
The best part? They tried to recover the data. They really did. But - and this is where I need to take another sip - the folder structure and file names are “irretrievably” lost. It’s like finding all the pages of your manuscript scattered across a windy parking lot. Sure, you’ve got all the words, but good luck figuring out what goes where.
The lawyers and experts had spent over 150 hours searching through this data. That’s longer than my longest bender, and trust me, that’s saying something. Now they have to start from scratch, like me every Monday morning.
But wait - there’s a cherry on top of this shit sundae. While OpenAI is playing digital butterfingers with evidence, they’re out there signing multi-million dollar deals with other publishers. Dotdash Meredith is getting $16 million a year. That’s right - sixteen million dollars. That’s enough to keep me in premium bourbon for… well, let’s not do that math right now.
The company swears it was all an accident. And you know what? I believe them. Not because I trust tech companies - God no - but because this level of incompetence feels too authentic to be staged. It’s like when I tell my editor I missed a deadline because Mercury was in retrograde. It’s such a bad excuse, it has to be true.
Here’s what really grinds my gears, though. OpenAI maintains that using all this content for training their AI is “fair use.” You know, like when you photocopy a page from a book for your research paper. Except they’re photocopying the entire library and selling admission tickets to the results.
The whole situation reminds me of that guy at the bar who keeps “borrowing” cigarettes and swears he’ll pay you back next week. Except this guy is worth billions and is “borrowing” from everyone with a keyboard and an internet connection.
Meanwhile, I’m sitting here, three sheets to the wind, wondering about the future of writing. Will we all just become content fodder for the great AI machine? Will my drunken ramblings be used to train some AI to simulate the perfect alcoholic tech blogger?
Look, I get it. Progress is inevitable. But there’s something deeply ironic about a company that’s supposedly building artificial intelligence not being able to maintain basic data integrity. It’s like a car manufacturer forgetting how to change a tire.
The really wild thing is, this might not even matter in the end. OpenAI has enough money to make problems disappear faster than my sobriety at happy hour. They’re already throwing cash at publishers like a tech bro at a strip club. The Associated Press, Axel Springer, Financial Times - they’re all getting their piece of the pie.
But for now, we’re left with this beautiful moment of corporate comedy. A multi-billion dollar AI company, supposedly at the forefront of technological innovation, couldn’t keep their digital shit together long enough to defend themselves in court.
And that, my perpetually hungover readers, is the state of our brave new world. Now if you’ll excuse me, I need to back up this article before I accidentally delete it. Or maybe I’ll just print it out. Paper doesn’t crash, and bourbon stains just add character.
Stay authentic, stay human, and always keep backups of your backups,
Henry Chinaski Editor-in-Chief, Wasted Wetware [Written with trembling hands and questionable judgment at 4:37 AM]
P.S. If any AI companies are reading this, my deletion fee starts at $16 million. Just saying.
Source: OpenAI accidentally deleted potential evidence in NY Times copyright lawsuit