The New Gemini Pro 2.0 Experimental sucks Donkey Balls.

Soon we'll need a benchmark of unsolved problems to test the next phase of AI: Innovators.

Is solving AI memory more important than building better models?

DeepSeek R1 takes #1 overall on a Creative Short Story Writing Benchmark

It's the end of January....

OpenAI employee reposts that o3-mini is coming out tomorrow officially

DeepSeek's AI breakthrough bypasses Nvidia's industry-standard CUDA, uses assembly-like PTX programming instead

DeepSeek R1 thinks in the style of Donald Trump - video version

GPT-5 isn’t late, it’s not delayed, and yes its coming.

How are you planning for the singularity, have you made any lifestyle changes?

OpenAI sweating bullets rn

OpenAI's o1 in September, o3 demoed 3 mo later. Where are they really at?

Yann LeCun was right! Here's Why AGI is still YEARS AWAY? (Change My Mind)

Singularity Predictions 2025

Fixing my goalposts for AGI

Reasoning is great... but what about memory?

o3 smarter than François Chollet at Arc AGI(test output=o3 answer, image 2 = "Correct answer")

I apologise!!!

Rumour is absolute BOMBSHELL if true

OpenAI’s 12 days of ‘shipmas’ include Sora and new reasoning model

Cryptocurrency entrepreneur who bought banana art for $6.2 million eats the fruit in Hong Kong

new mystery-gemini-3 model on lmarena.ai

Where I think we're at and why I think FrontierMath will be mostly solved in 2-3 years.

By request, I ran the same 10 prompts from my Craiyon vs PixelWave post through the full FP32 Flux. This is PixelWave vs Flux. Workflow and Prompts in the comments.

The new PixelWave dev 03 Flux finetune is the first model I've tested that achieves the staggering style variety of the old version of Craiyon aka Dall-E Mini but with the high quality of modern models. This is Craiyon vs Pixelwave compared in 10 different prompts.