{"id":7633,"date":"2026-05-15T16:58:04","date_gmt":"2026-05-15T14:58:04","guid":{"rendered":"https:\/\/amaelle.life\/?p=7633"},"modified":"2026-05-30T13:15:48","modified_gmt":"2026-05-30T11:15:48","slug":"jag-bad-ai-kla-mig-till-meta-gala-och-sa-har-gick-det","status":"publish","type":"post","link":"https:\/\/amaelle.life\/en\/jag-bad-ai-kla-mig-till-meta-gala-och-sa-har-gick-det\/","title":{"rendered":"I asked AI to dress me for the Meta Gala \u2014 and this is how it went"},"content":{"rendered":"

It started as an experiment. I wanted to see what happens when you give AI an image and an idea. The result? A champagne-colored dress with a cape, a Greek sunset in the background, and the text Meta Gala 2026<\/em> in the upper corner.<\/p>\n\n\n\n

Not bad for a few minutes of work.<\/p>\n\n\n\n

But what's interesting isn't the image itself. What's interesting is what's actually happening behind the scenes \u2014 and where this is going.<\/p>\n\n\n\n

Here's how it works, in short<\/h2>\n\n\n\n

AI imaging tools are trained on vast amounts of images from the internet. They learn patterns \u2014 how light falls, how fabric moves, what a face looks like at sunset. When you type a prompt, the model picks up those patterns and generates something new. Not a copy. A synthesis.<\/p>\n\n\n\n

There is no database of ready-made images for the tool to choose from. It creates each pixel from scratch, every time.<\/p>\n\n\n\n

That's why the results can be both astonishingly good \u2014 and sometimes scary wrong on details like hands and text.<\/p>\n\n\n\n

What tools are there and what is the difference?<\/h2>\n\n\n\n

It's easy to think that "AI imagery" is a single tool. It's not. Here are the most important ones:<\/p>\n\n\n\n

ChatGPT (with DALL\u00b7E)<\/strong>
What I used. Easiest to get started with if you already use ChatGPT. You write what you want in plain text, directly in the chat. Good for realistic images and portraits. Convenient for those who don't want to have a separate account on yet another platform.<\/p>\n\n\n\n

Midjourney<\/strong>
Used via Discord, which feels a little strange at first. But the quality \u2014 especially the atmosphere, lighting, and artistic images \u2014 is hard to beat. The favorite tool of many designers and photographers. Requires a paid account for serious use.<\/p>\n\n\n\n

Adobe Firefly<\/strong>
Integrated directly into Photoshop and the Adobe suite. A big advantage if you already work in the Adobe ecosystem. Trained on licensed material, which reduces copyright issues \u2014 more on that in a moment.<\/p>\n\n\n\n

Stable Diffusion<\/strong>
Open source. You can run it locally on your own computer if you have good enough hardware. Maximum control, maximum technical level. Not for the beginner, but powerful for those who want to dive deep.<\/p>\n\n\n\n

Canva AI<\/strong>
If you already design in Canva, image generation is built in. Not as powerful as the others, but enough for social media and quick projects.<\/p>\n\n\n\n

Personal account vs. work account \u2014 a difference that actually matters<\/h2>\n\n\n\n

Many people miss something important here.<\/p>\n\n\n\n

When you use a private account in, for example, ChatGPT or Midjourney, your images \u2014 and your prompts \u2014 can be used to train the model further. In most cases, this is stated in the terms of use, in small text.<\/p>\n\n\n\n

With a paid business account (ChatGPT Team or Enterprise, Midjourney Pro) you typically have:<\/p>\n\n\n\n

    \n
  • Better privacy protection \u2014 your data is not used for training<\/li>\n\n\n\n
  • Ability to keep images private instead of visible in a shared gallery<\/li>\n\n\n\n
  • Commercial rights to what you create<\/li>\n<\/ul>\n\n\n\n

    If you plan to use AI images in your business, on your website, or in marketing \u2014 invest in the right account. It's not expensive and the difference in what you can use the images for is significant.<\/p>\n\n\n\n

    Copyright \u2014 who owns the image?<\/h2>\n\n\n\n

    It's a question that no one really has a definitive answer to yet.<\/p>\n\n\n\n

    In Sweden and the EU, the starting point is that copyright requires a human creator. An image created solely by AI currently has no copyright protection \u2014 which means that technically anyone can use it.<\/p>\n\n\n\n

    But: if you as a human have contributed creative choices \u2014 specific prompts, editing, style and composition choices \u2014 it starts to get more complex. The legislation hasn't quite kept up with the technology, and case law is still evolving.<\/p>\n\n\n\n

    What you should consider practically:<\/p>\n\n\n\n

      \n
    • Check the terms and conditions of the tool you are using.<\/li>\n\n\n\n
    • Don't use AI images of real, named people without their consent<\/li>\n\n\n\n
    • If you sell products with AI images \u2014 read about commercial rights in your specific tool<\/li>\n<\/ul>\n\n\n\n

      Prompts that actually get results<\/h2>\n\n\n\n

      This is where most people give up too soon. They type \u201ca woman in a nice dress\u201d and are disappointed with the result. The trick is precision and context.<\/p>\n\n\n\n

      Try these and see what happens:<\/p>\n\n\n\n

      For portraits and fashion:<\/strong>
      \u201cElegant woman, 50s, platinum blonde hair, wearing a champagne embellished gown with a sheer cape, golden hour light, Greek island terrace, ultra-realistic, editorial photography style, shot on Hasselblad\u201d<\/em><\/p>\n\n\n\n

      For product images:<\/strong>
      \u201cLuxury leather bucket bag, lavender color, placed on a white marble surface, natural soft light from left, minimalist Scandinavian aesthetic, product photography, no shadows, 4K\u201d<\/em><\/p>\n\n\n\n

      For mood pictures for the blog:<\/strong>
      \u201cMorning coffee ritual, ceramic cup, linen fabric, fresh flowers, soft diffused light, top-down view, Scandinavian interior, muted tones, lifestyle photography\u201d<\/em><\/p>\n\n\n\n

      Notice the pattern: you describe subject, style, light, perspective, and technical feel. The more you tell, the more control you have over the result.<\/p>\n\n\n\n

      Where is this going?<\/h2>\n\n\n\n

      The short answer: longer than most people think, faster than most people are ready for.<\/p>\n\n\n\n

      Right now, AI can generate still images that are hard to distinguish from photographs. In a year, video will be on the same level. In three years, personalized, animated brand images will likely be something done in minutes rather than weeks.<\/p>\n\n\n\n

      It doesn't change the need for a human voice, a clear aesthetic, or a well-thought-out brand. It changes how quickly you can produce visual content that matches that voice.<\/p>\n\n\n\n

      Pros and cons \u2014 honestly<\/h2>\n\n\n\n

      What is genuinely good:<\/strong><\/p>\n\n\n\n

        \n
      • Speed. A picture that used to take a photographer half a day now takes five minutes.<\/li>\n\n\n\n
      • Accessibility. You don't need a budget for professional photos to have professional visual content.<\/li>\n\n\n\n
      • Iteration. You can test ten variations and choose the one that actually works.<\/li>\n\n\n\n
      • Creative freedom. Want to see your product in a context that's difficult to photograph? Just write it.<\/li>\n<\/ul>\n\n\n\n

        What still holds:<\/strong><\/p>\n\n\n\n

          \n
        • Details. Hands, text in images, and complex backgrounds are still weak points.<\/li>\n\n\n\n
        • Authenticity. An AI image of you is not you. It lacks the personal story that a real photo carries.<\/li>\n\n\n\n
        • Ethical gray areas. Images of real people without consent, distorted representation, billions of images without a clear originator \u2014 these are issues that the industry is still dealing with.<\/li>\n\n\n\n
        • Homogenization. When everyone uses the same tools with similar prompts, everything starts to look the same. Your aesthetic needs to be clear enough to stand out even with AI to help.<\/li>\n<\/ul>\n\n\n\n

          Final words<\/h2>\n\n\n\n

          AI imagery is a tool. Just like the camera was a tool, just like Photoshop was a tool. It doesn't replace creativity \u2014 it enhances it, if you know what I mean.<\/p>\n\n\n\n

          And sometimes it starts with an experiment. A dress. A sunset. And a text in the top corner that says Meta Gala 2026<\/em>.<\/p>\n\n\n\n

          What would you create if you knew it couldn't go wrong?<\/p>","protected":false},"excerpt":{"rendered":"

          It started as an experiment. I wanted to see what happens when you give AI an image and an idea. The result? A champagne-colored dress with a cape, a Greek sunset in the background, and the text Meta Gala 2026 in the upper corner. Not bad for a few minutes of work. But the interesting thing isn't the image itself. The interesting thing is...<\/p>","protected":false},"author":5196,"featured_media":7634,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_angie_page":false,"_kad_post_transparent":"","_kad_post_title":"","_kad_post_layout":"","_kad_post_sidebar_id":"","_kad_post_content_style":"","_kad_post_vertical_padding":"","_kad_post_feature":"","_kad_post_feature_position":"","_kad_post_header":false,"_kad_post_footer":false,"_kad_post_classname":"","_jetpack_newsletter_access":"","_jetpack_dont_email_post_to_subs":false,"_jetpack_newsletter_tier_id":0,"_jetpack_memberships_contains_paywalled_content":false,"page_builder":"","_jetpack_feature_clip_id":0,"_jetpack_memberships_contains_paid_content":false,"footnotes":"","jetpack_post_was_ever_published":false},"categories":[52],"tags":[],"class_list":["post-7633","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-tech-torsdag"],"jetpack_featured_media_url":"https:\/\/amaelle.life\/wp-content\/uploads\/2026\/05\/69AB7C76-92EC-438B-ADAE-D0C048B3FAD5.png","jetpack_sharing_enabled":true,"_links":{"self":[{"href":"https:\/\/amaelle.life\/en\/wp-json\/wp\/v2\/posts\/7633","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/amaelle.life\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/amaelle.life\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/amaelle.life\/en\/wp-json\/wp\/v2\/users\/5196"}],"replies":[{"embeddable":true,"href":"https:\/\/amaelle.life\/en\/wp-json\/wp\/v2\/comments?post=7633"}],"version-history":[{"count":2,"href":"https:\/\/amaelle.life\/en\/wp-json\/wp\/v2\/posts\/7633\/revisions"}],"predecessor-version":[{"id":7958,"href":"https:\/\/amaelle.life\/en\/wp-json\/wp\/v2\/posts\/7633\/revisions\/7958"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/amaelle.life\/en\/wp-json\/wp\/v2\/media\/7634"}],"wp:attachment":[{"href":"https:\/\/amaelle.life\/en\/wp-json\/wp\/v2\/media?parent=7633"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/amaelle.life\/en\/wp-json\/wp\/v2\/categories?post=7633"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/amaelle.life\/en\/wp-json\/wp\/v2\/tags?post=7633"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}