So, this isn’t true.
-
So, this isn’t true. It is not impossible to avoid AI-generated images in your presentations. But it is a bit difficult.
So I shall tell a little story.
My most recent talk is called The Mazy Web. It’s about poetry, why the web is important, what you can do to show your support, and open web advocacy. I’m pretty proud of it.
-
Stuart Langridgereplied to Stuart Langridge last edited by
I wanted a picture, a headline image, that would express a bunch of the things that I was trying to do with this talk. It’s about beauty, and ornateness, and weaving together an appreciation of language and literature with the need for us as a web community to come together in this modern world to stop those who want to suppress or destroy that for money. (If you think this is pretentious, just wait for the upcoming blog post about this.)
(2/n)
-
Stuart Langridgereplied to Stuart Langridge last edited by
So I envisaged an image which was sort of redolent of those themes. What came into my head was an ornate stone decorative circular wall piece, like something from the garden wall of a chateau, all made of curlicues and curves of stone, like the sort of thing you’d have making up the low wall dividing your stately home’s patio from the vast gardens. Capability Brown might have commissioned it. And then I’d alter the stone curlicues so that they spelled the name of the talk: The Mazy Web.
(3/n)
-
Stuart Langridgereplied to Stuart Langridge last edited by
And I did go looking for such a thing, and I couldn’t find it, not anything that matched the image that I had in my head. So I tried one of the AI art generators, and after some jiggery-pokery with the prompt I got a thing that was quite like what I was imagining. If looked at closely it was odd: a bit blurry and cartoonish. But it was fine. I scaled it up to a decent resolution using another AI thing, and edited it to have the words in a way I liked, and I made it the title image, thus.
(4/n)
-
Stuart Langridgereplied to Stuart Langridge last edited by
I did the first version of this talk at State of the Browser 2024 a free weeks ago with that image as the title. And then I spent every day feeling guilty and hypocritical about having used AI stuff to make this image. I’m not kidding: every day I thought about this and wanted to fix it. I was doing the talk again at Oggcamp a couple of weekends later, and so I resolved to fix it, because then it would stop making me sad about a thing I was otherwise proud of.
(5/n)
-
Stuart Langridgereplied to Stuart Langridge last edited by
I went back to my searches. Pexels, Pixabay, Wikimedia Commons, Flickr stuff for CC, and I couldn’t find anything for ages. But eventually I did. A 3d model called GARDENS OF SEMIRAMIS of a ceiling rose of all things. I customised it a fair amount, and redid the image, and made it the title image for the talk at Oggcamp, and it will be the title image going forward. And now I don’t feel it scratching at my head every day because I did a bad thing.
(6/n)
-
Stuart Langridgereplied to Stuart Langridge last edited by
It’s hard to find an image that matches what’s in your head. Stock photo libraries have woeful search, and are expensive. Places with the open source ethos (e.g., Pexels, Pixabay, Unsplash, the Noun Project) are amazing, and if you can find what you want there, take it and credit the creator.
The IDEA of AI art - describe my vision and get it - is fantastic. But it’s built by stealing the work of artists and redirecting their livelihood to already-billionaire techbros. Try to not use it.
(end)
