(I've been using this as a test prompt for years. This is the first time I haven't got slop back)
@fofrai
-
Fable invents a new color using eye lasers
By
–
I asked Fable to invent a new color, and I received my first 'paused cat'. That said, he decided to pursue a strategy of projecting lasers into your eyes to trigger otherwise impossible cone activations.
-
Gemini 3.5 Live Translate: spoken translation, multi-speaker, no Klingon
By
–
Gemini 3.5 Live translate: Stream in speech, and stream out the spoken translation.
— fofr (@fofrAI) 9 juin 2026
It also magically works with multiple speakers.
It does not work with Klingon (I tried).
Try it on AI Studio:https://t.co/pTVEPZ0llf
pic.twitter.com/FMtlj5G2gbGemini 3.5 Live Translate: Speech stream, and spoken translation stream. It also works magically with multiple speakers. It does not work with Klingon (I tried). Try it on AI Studio: https:// aistudio.google.com/live?model=gem ini-3.5-live-translate-preview …
-
Foffee agent helps make Gemma faster, proud parent feeling
By
–
I asked my foffee agent to help make Gemma faster. I felt like a proud parent.https://t.co/JuJLnRrPka https://t.co/X1aNZ3MGqf pic.twitter.com/vduCu0pSxr
— fofr (@fofrAI) 9 juin 2026I asked my foffee agent to help make Gemma faster. I felt like a proud parent. https://
huggingface.co/spaces/gemma-c
hallenge/gemma-dashboard
… -
Trying Omni image-to-video for anime intro with Krea K2 output
By
–
Trying Omni image-to-video with this Krea K2 output to try making an anime intro sequence. https://t.co/T8NglJXTQG pic.twitter.com/6dku0GGeY7
— fofr (@fofrAI) 5 juin 2026Trying Omni image-to-video with this Krea K2 output to try making an anime intro sequence.
-
Efficient CPU LoRA Trainer for SD 1.5 based on CompVis repo
By
–
Agent prompt for the lora task: > From first principles, in a new repo that can reference the compvis one, make an efficient lora trainer for SD 1.5 that works on a cpu. This is the original lora paper: https://
arxiv.org/html/2106.0968
5v2
…, this is the Stable Diffusion paper: -
Efficient CPU LoRA Trainer for SD 1.5 from First Principles
By
–
Agent prompt for the lora task: > From first principles, in a new repo that can reference the compvis one, make an efficient lora trainer for SD 1.5 that works on a cpu. This is the original lora paper: https://
arxiv.org/html/2106.0968
5v2
…, this is the Stable Diffusion paper: -
Testing Gemini 3.5 Flash and Antigravity CLI with Stable Diffusion 1.5
By
–
Today I'm experimenting with Gemini 3.5 Flash and the Antigravity CLI to see how fast and how autonomously the agents can do things. – It took 20 minutes to install and run the original CompVis Stable Diffusion 1.5 repo, get the weights, debug, run inference and generate an
-
Impressive long horizon agent tasks: What’s new and “woah” this week?
By
–
Where's the baseline for impressive long horizon agent tasks today? What are you seeing this week that makes you go "woah"?