so you’re saying that of I decoded these intermediate tokens I wouldn’t get coherent sentences, but rather something completely random that is just a covenient representation of the image, or perhaps some words that relate to the image (sth like “woman” “man” “marriage” “blonde” “dress” etc.)?
the worst part is that we have been able to do this long before LLMs, and achieve much better results. There exist filters that will cartoonify your image! deterministic filters - they always spit out the same image given the same input.
the problem with LLMs is that they go through this completely unnecessary level of indirection - they describe the image in text and then generate an image based on this text - essentially behaving like an extremely lossy compressor-decompressor pipe
is this what they’re using as a promotional graphic? these don’t look like the people in the photo at all
in the middle it says “the strongest vs the most terrifying” and they are homophones (both are read saikyou)
yeah, if I were a Lemmy dev, I wouldn’t take this proposal seriously because of this. I would be wondering if that person really knows what they’re proposing and if they’re planning to engage in the discussion themselves or let AI do it for them, and in the latter case it would just be a waste of time.
ever heard of pen and paper?
sadly there are people here on lemmy who say that unironically
modern day terrorists can’t be bothered with such extravagant, risky and expensive methods as hijacking a plane, they prefer less complicated and cheaper ones like driving into a crowd of people or a mass shooting
only by a third by 2050? that’s only 1.1% yearly inflation rate
it’s a metro, no need to move anyone…
she looks like misa from death note
yeah let me just give all of my code to elon for free
in case anyone’s wondering, according to data from the National Center for Health Statistics, in 2021, the life expectancy in Illinois was 77.1
I’d say chatgpt is more like a self-driving tesla stuck in huge traffic. you don’t have any control, it can break down easily, you’re moving slower than a bike, all the while thinking that people who chose the bike to avoid the traffic are losers.
tbh I don’t see anything wrong with using AI just to write the abstract, assuming the author redacts it afterwards. It becomes much more problematic if AI is used in the middle section of the paper, where it is crucial to present information as accurately as possible.
pro tip: try cooking for 2 days or more at once