To me there are different motivations to include imagery and some of them are just fine with AI generation. If the imagery actually goes to the substance of helping to illustrate or explain the actual content then it seems very superficial to reject it based on how it was generated. In fact it is kind of stupid to insist someone slavishly draw something by hand that could have been done an easier way.

The problem comes when the imagery is placed as a false value signal. If it is designed to send a signal of "this content took a lot of effort to produce because an image that would be hard for a human to make is attached to it" - but the image was generated without that effort, then it's basically a lie and insult to the reader.

It's like styrofoam stucco on the ceiling.

Decoration and ornament have served always to signal extra effort that went into something, signaling that it is considered important and valued by the maker/owner. But people also have incentive to do cheaper mimicry. And then the viewers have incentive to distinguish real from fake to see where true, scarce effort went. These dynamics are crucial for understanding the feelings and emotions around AI. Value is almost never in the thing itself or some inherent quality but in the social signals and relationships. (somewhat less true in very utilitarian engineering machinery or for sustenance level basic necessities).

The most important thing to keep in mind, when considering AI images for helping to illustrate or explain the actual content, is that AI-generated images have a big trust problem.

Take a look at the Google Image Search results for 'Engine Diagram' https://www.google.com/search?q=Engine+Diagram&tbm=isch

In the first row I get two correct-looking diagrams, then a load of AI-generated ones. Three cylinder heads for four cylinders? Two crankshafts? Oil pan labelled Crankshaft? Oil pan labelled Connect in rod? A four-cylinder engine with seven valve springs? Parts labelled up to 39 but a 9-item key? Spark plug wires that loop back on themselves? A crank shaft where the pistons aren't offset in phase? Four-cylinder engines whose exhaust manifolds have three or five openings? Then in the second row, I get "Parts of a V6 engine" showing a four-cylinder engine - with an 'Exhaust malvol', an 'extaust manivol', an 'exsoust manivol' an 'oilolaivt manivol'

Savvy technical audiences have noticed this - and so when a diagram has a whiff of AI about it, they become extremely skeptical and distracted.

Personally I only use AI-generated diagrams very sparingly, and very carefully.

>In fact it is kind of stupid to insist someone slavishly draw something by hand that could have been done an easier way.

Communicating low effort used to be insulting to the consumer, yet now I'm expected to praise it? I understand that we're not talking about monetary exchange, but we're still talking about my finite time in this world, time that I could've been invest reading or looking at something else.

So we're going back to my favorite adage these days: If you cannot spend time writing it, I refuse to spend time reading it.

In fact, the moment I realize someone's content is AI generated, I just feed the links (youtube -> transcript, blog posts -> link) to chatGPT and have it generate a summary. That's because if you're in favor of using your time efficiently then so am I.

The only reason most of my blog posts have an image at all is because social media seems to want some type of thumbnail image to go with the post. If I don't have something, many of them seem to just grab my logo instead which seems...spammier.

I try to find something better when I can, but the need to put something on every post is exhausting.

This. If it's actually an illustrative part of the text, include it, even if it's AI generated. Not everyone who can write can do graphics.

If it's only decoration, leave it out. No matter whether AI, stock or custom made.

True until LLMs become smarter than humans. At that point, you __want__ the AI generated image because it signals that AI was used.

> The problem comes when the imagery is placed as a false value signal. If it is designed to send a signal of "this content took a lot of effort to produce because an image that would be hard for a human to make is attached to it" - but the image was generated without that effort, then it's basically a lie and insult to the reader.

Well, the issue is the majority of AI images seems to fall into this category, which then trains my brain to treat all AI images in this way, rather than sifting for the few golden use cases where it's legitimately adding value.

At work, for example, I'm more likely to go brains off if someone presenting content shows an AI generated image in their slides.

You can call it superficial, but I’m not reading your AI blog, I’m not eating at your restaurant with AI food pictures, and I’m not listening to your AI music.

I’m quite happy without all the slop.

For every AI snob that's not reading an article because they see an AI image, there's 10 non-snobs who appreciate a decent usage.

If you're creating complete slop, no one is reading your article in the first place. Full stop. The image doesn't matter.

As with anything, it's about what you created, and creating at all - not the tool you used to create.

The snobs and haters will continue to b!tch and moan. They always have. Always will. Your two cent elitist opinions are still only worth two cents.

No. Even setting ethical considerations aside, AI-generated art is lazy, looks gross, and signals poor taste whilst wasting computing resources. I think it's a perfectly reasonable principle to claim, especially if we're discussing diagrams. I'd much rather look at a low-effort hand or MS Paint drawing, there's more humanness in that than any generative AI turd.

Same problem for me for sending emails, outreach, and other behaviors that signal human identity but it turns out to be an AI post. The area with grace for me is how much effort a human spent on composing the message - I wouldn't expect a message a human wrote >50% of and had an AI help grammar check, proof read and edit need an AI disclosure. But otherwise, if the identity is >50% have an AI disclose it.

This is how I'd like for human-on-behalf-of-human outreach to work too, i.e. if my human secretary sends a message under my identity and via my email client, I'd expect it to be under their name not mine.

This points out some long standing issues with lack of accountability and deception by humans through our time. Think: humans have been doing these things:

- human have other humans impersonate them and send messages on their behalf (i.e. via secretary, receptionist, email, a VA, via social media via a social media manager)

- friends or family answering your letters if your a celebrity

Those behaviors, deceptive and lies, are insults to you at the expense of the gain of the liar. We haven't held them to accountability or a standard of respect before AI, and so we are facing the same challenges we did before AI - we just have more humans doing the same thing with AI as the <insert old human role here>. Why have we been unable to hold them accountable to the deception before - lack if transparency, lack of power to hold accountability to the behavior?

Come on. We all know what kind of images this post is talking about. No need to play dumb.

> In fact it is kind of stupid to insist someone slavishly draw something by hand

I believe they used to call that "art."