The other night it regurgitated a humorous four-stanza poem for my D&D group, incorporating (and even rhyming) several of the player character names. Given that the names were made-up and unique, and have zero hits on google, which existing works would you say ChatGPT was regurgitating from?
Being able to answer this question will earn you millions, of not billions. However, considering neural networks works by banging their training sets together and emitting the resulting sparks, it certainly regurgitated its training set.
Writing a "4-line humourous poem" is also trivial. I'm sure my four year old could manage it. Praising a machine learning algorithm for this, that took over 10,000 Exaflops to perform its training routine seriously demeans human ability.
Also: it was a 4-stanza poem, three rhyming couplets per stanza. It had a recognizable beginning/middle/end and a reasonably humorous punchline. The rhyme scheme was AABBCC - your four-year old will need to know that when they replicate the feat ;)
4 stanzas. My bad. Still not impressed.
Comments like this add nothing to the conversation, please consider not posting them.
>Given that the names were made-up and unique, and have zero hits on google
You said it yourself. I just don't see the novelty that a machine learning algorithm, trained on huge amounts of stuff scraped off the internet, can find rhyming words.
That aside, there's an enormous amount of poetry in the training set. Given the relatively strict constraints of the style, there's not much interesting about slotting in words that grammatically make sense.
ChatGPT uses sub-word tokens. This is not a surprising outcome.