OpenAI’s cooking up something new – a large language model (LLM) supposedly gifted in creative writing. Sam Altman, the big cheese at OpenAI, is already singing its praises. But this AI’s debut short story has split the room faster than a dropped hard drive.
Is it a genuine leap forward? Or just another nail in the coffin for human writers, those folks OpenAI allegedly trained its models on without so much as a ‘by your leave’?
AI: Now With Feelings (Allegedly)
According to Altman’s post on X, this unnamed model is “really good” at creative writing. He even shared a 1,172-word metafictional short story it generated about AI and grief. Metafiction, grief, AI… sounds like the recipe for a prize-winning novel, or maybe just a very long afternoon.
TechCrunch notes that OpenAI hasn’t exactly been flexing its creative writing muscles before. This suggests they might think their new model is a significant upgrade in the word-slinging department. Which is good news for them, because as anyone who’s ever asked ChatGPT to write a haiku knows, AI hasn’t always been a literary powerhouse. More like a literary outhouse.
CNET posits that this move signals OpenAI’s ambition to conquer new territory beyond mere accuracy. Apparently, number-crunching and code-slinging weren’t raking in the dough as expected, so why not try fabricating emotions instead? Reece Hayden, a market research analyst, suggests OpenAI might face a bit of pushback from the creative industries, who are probably thrilled to see their intellectual property being used to potentially replace them. You know, the usual.
Critical Reception: From ‘Beautiful’ to ‘Doubly Trash’
The response to the AI-generated story has been… varied. Shocking, I know.
Novelist Jeanette Winterson, writing for The Guardian, finds the story “beautiful and moving”, praising its “understanding of its lack of understanding.” Which is either profound or deeply confusing. Regardless, she suggests it’s time for us humans to start reading AI, because apparently, AI is already reading us. Which sounds like the start of a Philip K. Dick novel, ironically.
Lance Ulanoff at TechRadar calls the story “remarkable” and unlike anything he’s ever read from an AI. Strong praise, indeed. He warns of a dark future for human authors, envisioning publishing houses churning out AI-generated epics that are “emotional, gripping and indistinguishable from those written by George RR Martin.” Let’s hope AI can handle complex character development and avoid plot holes the size of Westeros.
However, Kyle Barr at Gizmodo pulls no punches. The new model’s writing, he says, “still sucks.” Ouch. While OpenAI tinkers with updates, Barr believes they’re running out of steam. He accuses Altman of trying to flog ChatGPT subscriptions by promising uncreative people they can usurp the “literary ‘elite.’” The final verdict? “Even if you imagine a human created this, it’s still trash,” he writes, concluding, “Knowing AI created it makes it ‘doubly trash.’” Tell us how you really feel, Kyle.
The Bottom Line
Is this new OpenAI model a breakthrough? The jury’s still out. The reaction to the piece of generated writing is mixed – as to be expected. But there’s one thing for certain – it has generated discussion, and we wait with bated breath for future developments.
Ultimately, whether this AI becomes the next literary sensation or just another tech demo remains to be seen. But one thing is certain: the robots are coming for our jobs, one vaguely emotional short story at a time.
Leave a Reply