AI picture technology has been unbelievable for the final yr, nevertheless it was nonetheless extraordinarily tough to generate life like photos with particular topics, individuals, or merchandise. Google’s new AI mannequin can lastly do it, marking the top of the industrial images business.
As artists, we wish to faux that firms respect our work; that they’re prepared to pay a premium for our pictures due to some intrinsic inventive worth, however in actuality, it is merely a enterprise expense. Images, or imagery usually, is a small line merchandise on the advertising and marketing price range. Firms don’t care who took the image, if it was shot on movie, if it required advanced lighting, if it was Photoshopped, or if it was generated by AI, as a result of their clients do not care. The images merely have to catch a possible buyer’s consideration and persuade them to purchase.
The inventory images market was already decimated years in the past by microstock web sites that charged lower than $1 for royalty-free photos, however why pay a greenback for a photograph that has been used lots of of occasions when you’ll be able to generate a better-looking customized picture with AI?
Midjourney, my favourite AI picture generator, prices $30/month, however why spend that when each massive LLM like Chat GPT, Grok, or Gemini can even generate photos utterly free of charge? Inventory images was already nearly nugatory, however now it is utterly nugatory. Anybody on earth can now generate what would have been a world-class picture, immediately, and free of charge.
Till now, industrial photographers might nonetheless compete with AI as a result of firms wanted imagery of particular individuals, locations, or issues. AI fashions might create a generic picture of “headphones,” however they could not create “our headphones.” AI might generate a photograph of a “woman sporting a costume,” however they could not make her put on “this costume.” That’s till now.
Final week, Google launched its new AI picture technology mannequin, Nano Banana AKA Gemini 2.5 Flash. So far as I can inform, this mannequin is totally free. When you use it for hours like I did, you may run out of “tokens,” and they’re going to lock you out for a day, however if you happen to swap your Google account, you’ll be able to maintain producing.
This new mannequin can lastly do what we now have all been afraid of; it could possibly maintain topics constant picture to picture. You’ll be able to generate an individual, place, or factor from scratch after which make small adjustments to the picture with out altering the topic, or you’ll be able to add your individual photos, and Gemini will create new photorealistic photos utilizing an actual topic. This sounds too good to be true, however I have been taking part in with it for days, and it is unbelievable. Gemini generated all of those photos of my sizzling sauce model, Oliveum, in about 15 seconds after I uploaded an actual picture of the bottle and field.
Actually the one draw back is that these information look like locked at 2MP, which is ok for social media however not fairly sufficient for print. Midjourney can at the moment create 4MP photos, and I’ve run these information via AI upscaling software program, and I can get these information to look ok for print. In just a few months, Google will definitely unlock the next decision choice.
With this mannequin’s launch, AI has made nearly all industrial images out of date, and it’ll solely get higher from right here. It is miserable to have spent my whole grownup life engaged on a craft that seems to have turn into nugatory in a single day, however there isn’t a level in complaining or making an attempt to struggle in opposition to this; it is already occurred.
I’ll do a follow-up video about this and my predictions for the long run, however proper now I would like to listen to your ideas within the feedback beneath.
