Live data from Hacker News

My personal AI benchmark: “Generate an SVG of a frog with a Habsburg jaw”

frogs.vaguespac.es

91–100 of 101 posts

Re: My personal AI benchmark: “Generate an SVG of a frog with a Habsburg jaw”

#91
post #80

Earlier quoted context omitted.

I'm afraid you'll need to look up "Habsburg" yourself.

Nathan is right though. It specifies a Habsburg jaw, not that the frog is a Habsburg king/prince. It's understandable that the LLM will hallucinate a king from this, but it isn't what's being asked of it.

When I ask to create an image of a smartphone, it will most likely show the front and while it's turned on, even though it's not mentioned. For a person, it will add clothing.

Adding things that are not explicitly stated in the prompt but that are probable are the main benefit of AI, as far as I am concerned.

Re: My personal AI benchmark: “Generate an SVG of a frog with a Habsburg jaw”

#93
post #91
post #80

Earlier quoted context omitted.

Nathan is right though. It specifies a Habsburg jaw, not that the frog is a Habsburg king/prince. It's understandable that the LLM will hallucinate a king from this, but it isn't what's being asked of it.

When I ask to create an image of a smartphone, it will most likely show the front and while it's turned on, even though it's not mentioned. For a person, it will add clothing. Adding things that are not explicitly stated in the prompt but that are probable are the main benefit of AI, as far as I am concerned.

I'm afraid you'll need to look up "frog prince" yourself. It was fictional.

Re: My personal AI benchmark: “Generate an SVG of a frog with a Habsburg jaw”

#95
post #77

Earlier quoted context omitted.

> Two of them knew they were extrapolating ("because Habsburg") and did it anyway. You seem to imply that they ought not to. I disagree. I wasn't familiar with the term before this post. Having learned it, were I given the task, I think I'd be strongly tempted to do the same extrapolation. > If you're deciding which model to trust with instructions, "how much does it embellish beyond what I asked" and "does it behave…

> You seem to imply that they ought not to. I disagree. That would be like saying anyone with Lou Gehrig's disease must look like Lou Gehrig. So we'll have to agree to disagree here.

Well, no, I think you create a pretty clear false dichotomy there.

Certainly a person with Lou Gehrig's disease could look like Lou Gehrig; and especially in a cartoon illustration, where it's difficult to convey the point, this kind of artistic license is used specifically so that the viewer will make these kinds of associations.

I would, likely, otherwise perceive an underbite as just an underbite.

Re: My personal AI benchmark: “Generate an SVG of a frog with a Habsburg jaw”

#96
post #80

Earlier quoted context omitted.

I'm afraid you'll need to look up "Habsburg" yourself.

Nathan is right though. It specifies a Habsburg jaw, not that the frog is a Habsburg king/prince. It's understandable that the LLM will hallucinate a king from this, but it isn't what's being asked of it.

An artist given the same prompt would likely do something similar.

Re: My personal AI benchmark: “Generate an SVG of a frog with a Habsburg jaw”

#99
post #71
post #69

Earlier quoted context omitted.

rex paludis is a nice touch.

It's even Carolus rex paludis! Not sure if fable is a history buff or a Sabaton fan, but I repeat myself.

Frogando III of Haunchburg, AKA Frognando the Tadporrible

Re: My personal AI benchmark: “Generate an SVG of a frog with a Habsburg jaw”

#100
post #93
post #91

Earlier quoted context omitted.

When I ask to create an image of a smartphone, it will most likely show the front and while it's turned on, even though it's not mentioned. For a person, it will add clothing. Adding things that are not explicitly stated in the prompt but that are probable are the main benefit of AI, as far as I am concerned.

I'm afraid you'll need to look up "frog prince" yourself. It was fictional.

Not sure what the argument here is. I expect AI to consider fictional entities if they are semantically connected to the prompt.
Post reply on HN