After reading Despertaferres’ latest post I decided to have a play on ChatGPT with modifying some images and had a lot of fun. Note – Guru is a technological Luddite and anything I say in this post should be taken with the knowledge that you are reading the work of a technophobe!

Here is the original photo I am using:

This appeared in a recent post. Some ChatGPT changes

ChatGPT’s version of a Dystopian style – I like this one.

A good representation of a Monet style. Although I am a huge Monet fan this one just doesn’t cut it for me.

ChatGPT’s version of a Dali style. Not as askew as I thought and after a few tries this was the best Dali representation I could get – again not for me.

A Rembrandt style version that I quite like.

A John Constable style. Kinda quirky but I like it.

I then thought I would try the same image with Google Gemini.

Hmm. Matched my view of Google Gemini’s drawings.

Nano Banana is Google’s AI image model for Gemini, and IMHP is not a very powerful tool. OpenAI developed its own competing image model, GPT-Image-1.5, for ChatGPT. Both have moved from a text based model to a Multimodal one but ChatGPT seems to provide a richer result.

Thanks Despertaferres’ another afternoon successfully wasted!