I'm having major fomo about missing @aiDotEngineer in London with my GDM colleagues. So much good stuff happening there this week.
@fofrai
-
Creative Writing with Mythos AI Without Complex Prompting
By
–
Creative and unique writing from Mythos without complicated user prompting.
-
Prompt for AI-Generated Concave Mirror Pendulum Demonstration
By
–
Saving this prompt for later:
— fofr (@fofrAI) 6 avril 2026
> a scientific demonstration of a concave mirror, a face-on view of the mirror as a pendulum swings in front of it, towards and away from it, through the focal point of the mirror, the pendulum ball has a two color design https://t.co/Iq9956xElESaving this prompt for later: > a scientific demonstration of a concave mirror, a face-on view of the mirror as a pendulum swings in front of it, towards and away from it, through the focal point of the mirror, the pendulum ball has a two color design
-
Media Generation Still Far From Truly Solved Creative AI
By
–
We’re still so far away. Solved media gen would be imagination realised, it needs to be predictable and rule following and prompt adhering, but it also needs to be malleable and novel, with the capacity to surprise and invent, and go somewhere new. Current state of the art media
-
Tips for Prompting Lyria 3 Music Generation Model
By
–
Some initial tips for prompting Lyria 3: – prompt your musical style separately to your lyrics
– prompt for the subject of your lyrics, and set context for the lyrics for best results
– you don’t need to specify instruments, but if you want something unusual, it works, like a -
AI Music Generation API Produces Full Tracks With Lyrics
By
–
Generate full tracks, with lyrics, available with an API.
-
Open Source Text-to-Sample Music Model with Loopable Output
By
–
An open source text to sample music model, with samples you can easily loop.
— fofr (@fofrAI) 17 mars 2026
A really clever concept.
Now I need to vibe code a UI layer. https://t.co/ZhAB41RAsUAn open source text to sample music model, with samples you can easily loop. A really clever concept.
Now I need to vibe code a UI layer. -
Imagen 4 LoRA Generalizes Well to Animal Fur Rendering
By
–
Interestingly it seems to have also generalised well to fur on animals too. Here's an Imagen 4 cat on the left, and the lora on the right
-
Imagen 4 LoRA Outperforms in Facial Hair and Details
By
–
Does guys too. Imagen 4 on left, lora on right. It's doing very well on beards and stubble. One of the bugs with this lora is that it likes to give people ear piercings, still need to fix that.
-
Imagen 4 and Kontext LoRA improve hair rendering quality
By
–
This is an output from Imagen 4 on the left, and using the kontext lora on the right. The parting in the hair looks much more natural, and the vae artifacts have gone too.