The AI Creator Generation

Image my four year old created using Midjourney AI Art Generator

Apologies about sending this newsletter couple of days later than usual. I blame it on the start of the World Cup. England vastly outperforming expectations with a thumping win, followed by two major upsets two days in a row. Argentina and Germany both beaten by underdogs. This World Cup is a lot more exciting than I expected!

Ok, enough about the World Cup. Back to ordinary business.

The image of the sports car was done by my four year old in about 10 minutes using an AI image generator called Midjourney. I asked Quinn what image he wanted to make. He said 'sports car on track'. His eight year old brother added 'in France'. I suspect it's because he's been watching the Monaco GP on Netflix. 

After I typed in the prompt 'sports car racing on track in france, unreal engine, intricate', Midjourney took about a minute to generate four images. Quinn then had a choice of enhancing (U commands) or creating variations of (V commands) one of the four images. Quinn clicked on V4 to generate variations of the the bottom right image.

With four new variations out, Quinn decided to enhance the bottom right image this time round. After a couple more rounds of enhancement, Midjourney generated the image at the top of the newsletter. The entire process took less than ten minutes.

My six year old is into basketball. She asked me to type 'basketball player shooting'. I asked her whether she wanted a realistic image, or a Studio Ghibli style animated one. She chose realistic. I explained to her Unreal Engine is a real-time 3D animation tool. To generate an image in the realistic style she wanted and see more details in the image, we had to add 'unreal engine' and 'intricate' in the prompt.

Following the same process as Quinn's, Sienna generated the image above. She is very happy with it.

I am sharing this story because I think this is a powerful example of how human creativity augmented by technology is fundamentally changing the nature of work. 

If even 4 and 6 year olds (with some help from their old man) can create images as polished as the ones you see here, do we still need designers who move pixels around for a living? Probably not. Do we still need designers with the imagination to come up with newer, more powerful ways to communicate an idea? Absolutely.

Quinn and Sienna are using basic prompts like 'sports car racing' and 'basketball player shooting'. As they become more familiar with Midjourney, they will become more adventurous with the prompts. How about 'Two F1 sports cars night racing in Singapore', or 'Lebron James scoring the 38388th point of his career to become the all-time highest scoring NBA player'? What will Midjourney generate?

The human genius with an AI tool as powerful as Midjourney is in coming up with creative prompts to get the AI to generate images far beyond what the human mind can imagine, and figuring out how best to use the final images. An image without context is just another image. 

The image above, for example, is completely unremarkable. But in the context of an ad for a hair colouring product, it works perfectly. The first time Clairol asked this question in 1957, the answer was 1 to 15 -- as in, only 1 in 15 people were using artificial hair colour. Just 11 years later, the answer was 1 of 2, according to TIME Magazine.

I hope my Midjourney experiment with my kids shows how schools are light years behind where the world is at and what technology is capable of.

Instead of rote learning, schools should be where human imagination flourish.

Sadly, schools are doing quite the opposite.

If you have done fun experiments with your kids, would love to hear it!

If this was useful to you, please share it with your friends!     

Subscribe now and follow me on LinkedIn, Instagram and Twitter for thoughts on the future of work and learning.  

Till the next issue!