rafabulsing

Yeah, that's exactly my though every time this argument is brought up. "Obviously this endeavor is not economically viable if I had to follow the rules, so I get to skip that part" is such an astounding line of reasoning.

I don't really have anything to say other than the fact that you both don't thing there's a difference between using AI and actually creating something yourself, and that you also don't care for explaining why those are the same thing to you, but you also still feel the need to nevertheless reply anyway... It's fitting, is all.

The thing, which anyone who's ever done any type of art realizes, is that 99% of the time, there isn't really such a thing as a high fidelity image in your mind's eye.

You feel like there is... But the minute you try to bring it into creation, you find out just how absurdly fuzzy it really is. A non-artist might believe that the failure of bringing your mind image to life is a failure of _skill_. That you are just physically incapable of accurately reproducing the thing you're picturing in your head.

While that might well be a factor, it's never the whole story.

That's most clear with writing. Anyone who's ever written anything of substance has had the experience of having what feels like a pretty solid story/thesis/whatever. But the minute you try actually writing, you realize just how fuzzy that idea really was. You have to work and labor at it to make it become something coherent. It's not a limitation in mechanical skill: typing absolute non-sense is exactly the same difficulty as typing the a NYT best seller. The difficulty is in the deciding what to type.

A huge portion of art is in that part of the process. Exploring ways to bring into existence not _the image you had in your mind_, because there's no such thing. But to bring into existence something that makes you _feel_ the same way the fuzzy idea in your head made you feel.

And that's the thing that AI does not, and fundamentally cannot, give you. All the innumerable micro choices your mind does as you try to gradient descend that raw log of wood into the carving of a bear.

Looking at an AI output and deciding whether it happens to match the fuzzy idea you had enough or not, simply isn't the same thing. I'm not even saying that it's not valuable or useful, it certainly can be. But it's fundamentally not _art_. That's the reason Rick Rubin, for all the praise he (deservedly!) gets for all his work, does not describe himself an _artist_. It's just not the thing he is doing. He doesn't create, he evaluates. And if you take that for what it is, that's absolutely fine, great even, when you're as good as he is.

But if you believe the evaluating is, as you put, "seeing what's in your mind's eye at the highest fidelity possible", that's simply untrue. And it _cannot_ be true, because there isn't actually a thing that's in your mind's eye in a way that would allow for that, regardless of whether you're using AI or creating it by hand. Using AI just gives you the illusion that there ever was such a thing.

That's true, but the very feeling that I might be inputting things in a way that might have to be updated later kinda drained my motivation.

I do understand that even if does get replaced eventually, it still was useful for a while, and probably served as the basis for whatever came after.

But still. For whatever reason, in my brain, it just diminished whatever psychological reward I got from adding information to OSM, which is a shame.

It doesn't help that there are multiple ways to map sidewalks in OSM (as a property of the roads themselves, or as standalone paths parallel to the roads).

I got pretty into OSM for a while and that's definitely the thing that eventually put me off... So much stuff has multiple ways to be mapped, and it gets confusing/overwhelming. I understand that OSM has a very malleable data model, and that definitely has many benefits, especially for niche/edge cases. But I do wish the community had arrived at a settled convention for at least the most basic stuff, like the aforementioned sidewalks.

Same! Somehow I had never heard of this. Will have to try and visit someday!

Also, lol @ the name of the river near that site. (For the gringos: "Rego Grande", "rego" officially means ditch/trench, but nowadays it's much more commonly used as slang for butt crack. So, it's the Large Butt Crack river. Which I guess tracks for Brazil :P)

Moises is able to do that! Though it's unfortunately not open source. Result quality can vary as with anything AI, but ime it usually performs ok.

Respectfully, his graph does not need 3 dimensions because it's a personal spreadsheet he uses just for his own training, so he can just display the data for his exact body weight.

Yeah, that makes a lot of sense!

Hah, that's very interesting to learn. I've always thought the little flag was lifted by the mailman, so that the homeowner could tell from a distance that they got mail. Cool that it's actually the other way around.

What exactly is wrong about a solution that perfectly fits the most common use case out there, at a fraction of the cost/complexity?