Encouraging signs of fitness progress with my morning bike rides: "Compared to a matching 5.8-mile ride in late September, you held the same time and speed with lower cardiovascular strain." I'll have to ride even harder, lol.

I read 2 @xdadevelopers articles today about local AI models for home labs. Key numbers: 744B. 120B. 27B. The biggest number doesn't align with which model could finish my work. Beast’s results: 24GB RTX 3090, ~203K context, Hermes agent shipped a good ComfyUI workflow.

Living in Central Florida today, it can be easy to forget that you're in the Deep South: the prejudices that once made it the land of segregation, lynchings and other forms of hate still simmer just below the surface for far too many people.

Today, my son and I got into the car and he said: "I want to play some music we haven't listened to together before that I found the other day." Then he played the Stone Temple Pilots album Core. I love that he's independently discovering a taste for the music of my youth!

To this day, uncanny aliens – my genesis collection of 29 1/1s on Foundation, now preserved on @networkeddotart – remains among the work I’m proudest of.

BIP 🧵 My MoE AI model was stuck in Hermes. I went model shopping and found an unexpected VRAM bottleneck: I'd been too cautious with it. The GPU must fit both the model and its working memory. My 24 GB RTX 3090 now runs a 27B model with 180k context.

few things take as long as setting up every single app when you switch phones the whole 'transfer device' setup thing isn't nearly good enough yet

folks aren't buying my art most of which isn't listed for sale there's clearly a conspiracy

I often have a couple fragments of the lyrics to "Once in a Lifetime" by The Talking Heads pop into my head. I have taken to altering and adding slightly to them, as an affirmation: This IS my beautiful house. This IS my beautiful wife. This IS my beautiful life. Well, Igor... That IS your beautiful house. That ... Show more

this is me I am my art

unhealthy levels of stress are unhealthy that's all I have to say about that

I raised Beast’s GPU cap 36%. Decode improved just 6–8%, while loaded UPS draw rose ~60W. Moving two layers’ MoE experts to CPU helped make room for 50% more context. Memory tuning paid off. These are live Kanban results, not matched benchmarks.

Local LLM/Hermes folks: proven better model on 5950X, 3090 24GB, 64GB RAM than Qwen3.8-Flash-Next UD-Q2_K_XL (65K, n-cpu-moe=32)? Production metrics: 28.1 decode / 244.9 prompt tok/s over 1.59M gen tokens; second window avg: 33.7 / 255.2 140W = current GPU thermal cap