The generative and creative AI explosion is here to stay.
The prompt to image hype of mid 2022, driven by the likes of Midjourney, was not just a hype after all.
It was a first step in a bullet-train race which reached the conversational AI station in late 2022.
Now the train has split into a set of gigantic pinball multiballs, causing havoc across all industries.
“Knock - Knock”
“Who’s there?”
“Me, a human”
“Welcome human. What’s your prompt? Think carefully - the few words you say will determine your fate”.
I stumbled the other day on promptperfect, a SaaS website that offers to optimize one’s prompt, taking into account the specifics of the various models that could be used. This works, allegedly, for chat-like LLMs, as well as for prompt to image applications. Prompt engineering has been a thing for a while, but I have never really dived into it - yet. Intrigued, I decided to give it a try - on the image side - to create a couple of pieced for my Insta profile, keeping up with an exercise initiated last year for #CreativeAI_Sydney, emulating famous rock album covers.
This is a follow-up from my earlier post.
This is really entertaining. It is also depressingly addictive. Self-indulging as well: what a luxury it is to be able to engage in a polite, seemingly rationale discussion with s̶o̶m̶e̶o̶n̶e̶ ̶w̶h̶o̶ something that can reply almost instantly to each point, yet can wait with an infinite patience for my responses.
It also is a great support for fictional Socratic dialogues(*) - a very effective medium to convey philosophical ideas by guiding a student to reach the conclusion through their own reasoning. I used it with ChatGPT to talk about “regulatory narrow focus” which is an issue I have been looking in for a very long time.
Earlier this year, Diapason participated in the organisation of the CreativeAI_Sydney symposium. At this time, public attention on generative AI was growing exponentially, and focused very much on Dall-E2, Midjourney and Stable Diffusion (in the order they entered the spotlight), i.e. ‘prompt to image applications’. Most of our discussions gravitated around the this fundamental question: “Can AI be an artist (and therefore be a threat to existing artists) - or is AI only a tool that artists can use”. This question had been answered 5 years ago almost to the day in in the field of medical imagery AI with the idiom Rads who use AI will replace rads who don’t, although this is still open to interpretation.
The invention and spread of writing did not turn everyone into a literary creator. Neither did the printing press, typewriters, home printers or word processors. The democratisation of home studios, electronic music equipment and pro-quality plugins did not turn every bedroom muso or DJ into Jean Michel Jarre or Deadmau5. Like Jordan Tanner @jrdntnnr explained to John Flowers, “having the latest, flashiest Nikon camera […] doesn’t automatically make you a professional photographer”.
Diapason’s resonances this past month
I just received my invitation to DALL-E 2 and have started playing with it as you can guess from the images on this post. I haven’t used Midjourney and Gaugan yet. Beyond the “fun” of it, I have been blown away by the quality and depth of some of some of the artworks I have seen recently, like this journey or this landscape.
I am so looking forward to explore this field further at the upcoming Creative AI Symposium. Prof. O. Bown has designed an exciting program with many talented artists, researchers and practitioners over 2+ days here in Sydney (and online). Free registration here. I am honored to chair one of the industry panels where we will hear more from Tomasz Bednarz, Johna Barthelemey, Bhautik Joshi and Jess Edwards
At Diapason we are really excited to be part of the organising committee for the upcoming Sydney Creative AI Symposium. This will happen in August 18th-20th at UNSW School of Art & Design, Paddington Campus, on Gadigal Land.
This event will bring together technologists, artists, arts organisations and researchers based in and around Sydney to showcase work, to take stock of what impact AI and generative technologies are having, now and in the future, to solve technological problems and to design the best world we can for arts communities. We are calling out to artists, entrepreneurs, cyber-musicians and AI whisperers to let us know what you are experimenting. Your ideas will shape the format of the event. The full annoucement is there.
Copyright (c) 2022, Diapason Consulting; all rights reserved. No cookie enabled: https://gohugo.io/about/hugo-and-gdpr | https://www.goatcounter.com/help/gdpr
Template by Bootstrapious. Ported to Hugo by DevCows.
