At least the images are easy to scroll past. Useless AI features that actively get in the way of real functionality are much more off-puttting to me.
I also hate them because when one of those images catches my eye, I feel like I've been scammed. It's like biting into a delicious-looking chocolate bar that's filled with ash instead of caramel and peanuts you reasonably expected. After enough of that, you'll even sensibly get wary of real chocolate bars.
I hope that gives you at least a slightly interesting angle on it?
Your other post is very spot on and the fact that you reflected on it gives you huge credit.
My thought was... as a computer/techie person, I have to try this and get with the programme. And to be honest I love much of what generative technology can do, and glad that I used it to find out. But I'm amazed how quickly I tired of it.
But while those deeper reflections on what and who I am aligning myself with are very important I ultimately caved to perceptions in a cowardly way rather than take a deeply principled stand. Still, it was one of those "Are We the Baddies?" moments noticing the skulls on our hats... :)
> does it really cheapen it to have an obvious AI thumbnail?
> Does it say "trying too hard"?
If the article isn't about AI images, they often signal one more of the things WordArt used to:
1. Cluelessness (using $HOT_NEW_THING out of FOMO)
2. Sales pitches (trying to sell you something)
3. Irrelevance (lead-ins for B2B sales to 1 and 2)
4. Hard social & technical problems
The last case is the worst:
1. They're about to discuss font rendering problems
2. Consider leaving before someone assigns you the ticket(s) for them[2]
[1]: https://filestore.community.support.microsoft.com/api/images...
To me it's a signal the author either doesn't care about the overall quality of their post, or doesn't know or care that it's no longer special.
Which may be fine for some kinds of posts but not a good look for anything significant.
For me personally, I tend to notice and obsess over an image's details when I see something's off. When AI art catches my attention this way, I end up going over it for a few minutes, which detracts from the time I have to read the actual article. In some cases, I just decide I've spent too much time and close the tab.
The image should be there to support your article, not distract from it. If you're going to use AI art, make sure it's indistinguishable from non-AI art.
I'll admit, I feel a certain cruelty to using AI-generated imagery alongside writing or information: it's disrespectful to the audience, and it cheapens the work.
that's ai generated content main trait
I saw Postgres post with a picture of elephants from Africa, Docker post with a picture of a whale in ocean. Like, why do you even want to waste your website's bandwidth for that? Nowadays there are too many of them that my brain just automatically ignores them.
I'm sure there are plenty of good uses of AI art in this space too (I think I've seen one or two creative AI images that worked well with the content), but the bad examples are overwhelming in number, and they stand out.
AI images are often at least as relevant as a stock photo, if not more so (assuming the creator spent the extra 30 seconds to make a prompt that's more tailored to the content)
That's damning with faint praise, as stock photos are rarely very relevant or add to the writing they're tacked onto.
I'm especially confused about this for feature images, because I would expect in most cases any helpful image (like a graph or diagram) would make for a really unattractive or incomprehensible thumbnail, and thumbnails are where featured images get the most exposure (social media previews etc).
I don't get upset when someone uses Bootstrap to do their site layout or React to handle interaction, despite the fact that a designer didn't get paid or a programmer wasn't hired to hand-roll the code.
I'm not affiliated with Kagi, apart from being a subscriber.
I wouldn't want to see images just auto generated in the next.js markdown to html build step from the post's tags (like stock images used to be summoned code-monitor-out-of-focus.jpg) though I'm sure that's going to be a thing.
besides, all those shitty blogs were using shitty stock images and bland twitter memes. budding artists generate revenue primarily by drawing risque pictures (primarily of someone else's copyright protected characters, but no one likes when this is brought up)
When the generative AI boom started, I thought it would be fun to use it to generate my next book cover. But I changed my mind some time ago. Gotta have the humanity in there, and AI just feels cheap, somehow.
[0]https://venturebeat.com/ai/science-journal-retracts-peer-rev...
1) The negative associations of readers mentioned by OP. People do assume that because some AI is used, all of the content is suspect by proximity.
2) The quality just ain't that good. Although we deliberately chose to limit to cartoon/non-photo-real, after a while you notice an ineffable sameness about all the AI stuff.
3) It takes away the fun of 5 minutes in Photoshop/Gimp
4) Maybe this has changed, or got worse, but we got a creepy sense of sinister censorship and embedded values in the AI, especially around race and gender bias. I got really tired of finding Gordon Freeman lurking in almost every image, whether prompted for or not.
5) Which raises the thought... maybe we are just shit at using AI image gen? In which case, why not use the tools we are good at (digital image tools) and have spent 20 years mastering.
The whole AI image episode reminds me of something I saw when working as a music producer. In the middle of an album, a band took receipt of a cool new drum machine (actually the tech companies just give expensive new toys to big bands and names) and it caused problems. It wasn't that the machine was "bad", it was just so cool and distracting that it sucked the creative energy out of the band and became the focus. Eventually the tail started to wag the dog and some members even started to ask "How are we going to use the MPC in this track?", as if it had become necessary.
That was the 1990s but I've experienced the same lesson about tech many times since. It's great to use new stuff. You should jump-in and know what is all about. But there's a time to walk on past the latest novelty and hype and not get hung up on it.
The music world is full of this sort of badness. The same thing happened with the vocoder, autotune, and so much more. Until people learned how to use the tools, the use of the tools made the music much worse.
I think the same effect threatens to happen with the use of generative AI for artistic purposes. The results of that use are going to be pretty bad until everyone figures out when and how it should (and should not) be used.
And I reflectm on my own biases, thus this observation.
Not saying that's right or wrong, just saying that it is.
I'm a little surprised it has become so common; it can't possibly _add_ value.
People who can't stand AI generated pictures and get triggered by the very idea of it are going to rush in here to complain about it, people who are like "meh, it's fine" aren't even gonna bother loading the post.
Could also have 矽 to indicate that you think AI is just a "silly con".