Anyways, without reference to anyone in particular, there’s a strong all-around need for proportionality in the responses to certain stimuli here in this thread.
(boy does replying to this feel like a waste of time)
What are you even talking about… There’s no forum on the planet where individual forum posts are translated… this has to be automated. Do you seriously think I don’t use various translation services?!
There’s a difference between small translations for specific applications that thousands of people will use vs forum posts that die after a couple weeks. The amount of text is also insane difference.
The fact you’re even trying to make this argument… honestly I don’t even know where to begin to argue because it’s just so … I can’t even…
NO… OF COURSE NOT. Are you reading my replies?
If you REALLY need to explain to you how … any art… AI or human… of completely irrelevant half naked mice holding up barely related things for this project… would be ok either way…
Then you’re a lost cause honestly…
Art should have been commissioned, either for free or paid by someone. BUT ACTUALLY RELEVANT. Something with a f… map on it atleast…
If that image of those half naked rats didn’t say JOSM, you wouldn’t even come close to guessing what it’s purpose was.
…
Honestly with the type of replies you make… I’m going to stop responding, because it’s LITERALLY a waste of time and electricity to type out responses to your replies.
For anyone who doesn’t understand why OSM mappers would be unhappy about these AI images on the JOSM startup page, let me attempt an explanation (however futile this may be). It’s much more nuanced than human art = good, AI art = bad. JOSM is a tool for humans to contribute data to a project that values human created and human curated data. AI generated anything is the antithesis of this. A person who opens JOSM is about to give away some human time and energy making a contribution to OSM. They are “welcomed” by an image suggesting “You aren’t contributing enough. You should also be helping to translate JOSM”. This is already off-putting, but the fact that the image has been generated by AI adds another layer of insult. The message from the JOSM maintainer(s) seems to be “Our time is very valuable so we will use AI as a shortcut while requesting that you spend more of your time contributing than you had already been planning on”. It’s clear that wasn’t the intended message, but if you can’t see how it comes across as a slap in the face I don’t know what else to say.
No, I didn’t think that, but I also didn’t want to presume on your thoughts (as that might’ve come off as rude).
What I wanted is to draw a parallel, if I may: “do you seriously believe that there is an working advertisement on a planet in 2025 in which automated machine tools aren’t producing the majority of the work on visual results of the picture?” Because I don’t.
Click to see boring details WHY I don't think so.
I don’t know if you’re ever used Photoshop or even a Gimp or looked in their source code. Even using simple effects like blur or auto-levelling is a primitive AI. As is reverse image search, stock photo smart-matching, etc. All the editing and processing and filters. God, the filters. All of that is already mostly machine-made, where the human editor only provides basic ideas (blur this part, enhance this part, cut this and replace with that, merge in this way etc). Do you see where I am going? Writing prompts for AI is not all that different.
Remember the Deckard from the original Bladerunner instructing computer “enhance 34 to 36, 25 left, stop.”? Those are AI prompts. As is firing up photoshop and selecting that drifter motion blur effect (or whatever) on the selection. UI is little different, but still, it is the machine which does most of the work, and human only directs it toward the wanted direction.
To explain the history: originally, in professional photography it was considered art. Having to set up the scene, frame, to produce right lightning and reflectivity, color balance, exposure, objectives, shutter speed, aperture, depth of field, filters, and million other settings – and later develop the film (with all its tweaks!) – when the human sets them all manually to the perfection, you got an art of photography. You may disagree, but it was an real art.
Then came point-and-shoot cameras which made part of that automatic (S,A,DOF,focus). Then came digital cameras which made more of it automatic (auto-ISO, anti-red-eye, HDR …). Then came iPhones and Androids with smoothness, vivid colors, picture enhance, beautify, filters etc. And all that machine preprocessing is in “original” JPG. Which then gets even more postprocessing in creation of the final picture in Gimp or Photoshop or whatever.
In any of those, vast majority of the work was done by machine under relatively little human instructions. And the trend is toward even less human instructions. Was most of the photography created 10 years ago still art? There are quite a few old folk who disagree. They say that you can’t make art with all those digital gadgets, and that art of photography died with era of silver halide and darkrooms. Yet they are some who say computer software is just “helping” humans be more efficient, and that is the fact the human is directing the computer is what matters for it to remain being “art”.
Yet, “art” is poorly defined concept. Because, if we accept that definition, then AI being directed by human prompts is still “art”, and many (including myself) are wary of taking that definition to its natural conclusion.
And a few that might be doing noticeable part by hand are going to disappear in next few years, because #profits… Thus my point is, that human drive for convenience is unstoppable (if we ignore things like civilization collapse, which would prove me wrong).
Click for depressing statistics
the books people read, are 99.9% mass-copied by practically any individual with a simple click, and not few learned in the art of Scribing
the pottery 99% of the planet buys is no longer hand-made art, but mass produced copies without soul for 0.99$ apiece
the eloquent human translations are a rarity today, existing maybe in 2% of the cases (mostly because of lawyers), and all the rest of it is done by sloppy machinery with quite bad (but passable for majority of simple cases, as you note) results
the real art of photography is today already down at 3%, with 97% being machine spewing out output with minimal artist involvement for maximum efficiency. And it is going to take another big dip as more tools become usable by people who have not spent decades in learning and practicing (but instead have just installed an “app” to do it automatically taking just minimal input).
5 year ago, the music was there at maybe 5% art, and rest machine generated, but it has taken a big dive. Have you looked at the progress there lately? (I know, you hate it, but “know your enemy” and all that. It is frightening how convincing fakes it can create. Getting more convincing than the real thing, so we’re about to run past the 50-50 point of the Turing test real soon now, eh?)
What is my point? Either people will change a definition of what word “art” means, or the “art” is going to be practiced by significantly less people (like in middle ages, when only those rich enough could afford to spend time and money on learning and practicing it). It certainly will stop being a money generator for vast majority (but not all) artist, just like most translators (or handmade pottery makers) can’t count on you or me to support them by paying them to translate each post (or to make a set of plates), so their numbers are dwindling too.
The best we can hope is what happened to professional chess. But I think it will be worse, as most people were not chess players, and most people do take pictures with their phone.
Do I think it is good future to be heading towards? No, I don’t believe so. Is it enough to convert me to a Luddite? Not yet, but I feel the appeal of it drawing ever closer (helped with other #enshittification of the tech)
Yes, I am. (I just wanted to make sure, and not put words in your mouth.)
Yes. So, to split the problem into component parts, why I think people are having issue here:
some hate AI, and anything that uses AI for any purpose whatsoever riles them up to the skies. (Because it kills human creativity, loses artists jobs, etc.)
some are fine with images of “fun cute animals” ads, but heavily disagree JOSM maintainer for his taste in what “fun cute animals” look like (but would be totally fine with, say, LOLCATZ instead of half-naked-threatening-looking-rats). (i.e. NSFW crowd)
some dislike JOSM maintainer for his stubborness to change his opinion, even when they try yelling very loudly to force him to adopt their opinions instead (the strategy that rarely works IME).
some dislike JOSM maintainer for the idea of putting up visual ads of any kind, especially changing ones that capture attention (i.e. anti-brainhacking crowd)
some don’t have a problem with AI images, but aren’t particularly impressed with results of this one
some dislike JOSM maintainer for preexisting personal reasons
some don’t like useless images needlessly wasting their (pay-as-you-go) bandwidth
some dislike JOSM maintainer for being greedy and expecting more help from the users of his software (to the point of being rude).
Did I miss anything? I’m mostly (2), (4), (7), (8). I think for most people it is basically just (2). Although for some it seems to be predominantly (1).
C’mon - whoever wrote that wikipedia article has got to be taking the proverbial. I mean, seriously “… it is still used as an example of principles such as Bayesian probability and implicit religion” . Did Alan Sokal write it?
I apologise for the lack of AI mice in this post. Here’s a real one**:
Some are also like reluctant to change. Like that start page of jOSM was for 20 years (or something like this) the place, where you found latest news of jOSM and only that.
That is indeed unfavorable interpretation, but to me those images didn’t came off in that meaning (nor do I think they would by default, unless I was already predisposed against person asking for help (“assume good faith” and all that).
I mean, when I see the post from OSMF asking for donations or becoming a member, I do not immediately think “you bloodsuckers, I already contribute so much time to the map, and now you want my hard-earned money too?!” .
Nor when there was a need for category moderator did I think “how dare you asking for more of my free time, is it not enough that I map all that stuff?”.
Nor when I contribute code PR for some feature in some OSM software do I default to “how dare they ask for changes to my beautiful code, after all the time and effort I put into that PR; they just don’t appreciate my work enough!”
Dunno, perhaps I am too hard-skinned to be riled up by such minor stuff, or I miss too much of the nuances of the English language, but to me all that rage seem waay too excessive / blown out of proportion.
It seemed to me like those were merely a dozen or so AI generated pics of dubious quality / appropriateness (and which can be turned off in settings), yes? Maybe with some sense of overentitlement (something which seems rampant nowadays)?
And yet people seemed as upset as with Trump’s renaming of Gulf of Mexico – which might have infinitely bigger consequences (and we don’t even have a setting to turn off Trump !)
True. I get that (and I’m as much a creature of habit as a next folk!)
Was it, though? There were calls for translation help as well as AI audio on that MOTD page a year ago too, yet no calls for revolution (that I remember). Or two or 15 years ago - they all were calling for “Help translating JOSM into your language” all the time, so that part haven’t really changed.
It is just that before is it was all hyperlink text, so much more ignorable than in-your-face inline images (which, as I gather, was exactly a reason for putting them up).
So I’d class that as (4) above (i.e. “dislike […] the idea of putting up visual ads of any kind”)?
Unless this is a euphemism I suspect this will be of no use. I think I heard recently that even snails tossed over the garden fence have been known to return successfully. I doubt a mouse would have much trouble.
It depends on the context. I made this very point about iD adding the “free to use but not free to run” slogan to its post-save screen. Fortunately, we haven’t heard of anyone getting turned off by the message, but the mitigating factor is that it only appears after you’ve saved your edits.
Now fixed, of course. Anyways, OpenHistoricalMap’s fork replaces the post-save screen’s call to action with a more positive message that OSM’s version can switch to easily if necessary.
This is evident in the discussion, but this also risks caricaturing a more nuanced critique, which is that the use of generative AI for this particular call to action smacks of irony.
(apologies for continuing the offtopic diversion, but) Not a euphemism - 3 mice were moved on that occasion, and each was somewhat characteristic in behaviour**, so I don’t think that those did return. There were 4 houses nearer the release point than mine, hence my comment. One evaded capture but wasn’t heard after Storm Éowyn. Obviously mice breed and other mice move back in to a vacant area pretty quickly, so other mice did come back… Re snails, it’s one of the more famous “crowd science” experiments in the UK.
** the one pictured was more Penfold than Dangermouse - caught because it was hiding under something on the floor but left its tail hanging out.