How the Fox with the Long Tail Learned to Play in the Dark Forest
I. Everything is a Balinese Cockfight It was after a police chase and a conspiratorial lie that Clifford Geertz and his wife Hildred were accepted into the social fabric of 1958 Bali. Until then they had been ignored, a treatment…
Themes?? What do I do with all this DHTML knowledge??
congratulations to rogue bus guy on the new job
“I am an argument against this age.
[…]
And I shall remain, for my own self hedges me in.”
- Adonis
I suppose there can be poetry of the Thicket
Writing is thinking
"On the value of human-generated scientific writing in the age of large-language models."
www.nature.com/articles/s44...
Voltaire’s garden has been paved over by the Achievement Society. Today, "cultivating your garden" just means documenting your burnout for the digital gaze. To stay human in 2025, you have to leave the manicured rows behind and head for the woods. The Deep Dark Terroir of the Soul
Voltaire’s garden has been paved over by the Achievement Society. Today, "cultivating your garden" just means documenting your burnout for the digital gaze. To stay human in 2025, you have to leave the manicured rows behind and head for the woods. The Deep Dark Terroir of the Soul
If the future is a terraformed plain, the only radical act left is to become a cultivator of shadows.
If the future is a terraformed plain, the only radical act left is to become a cultivator of shadows.
The case for the World Wild Web: We are trading the "thick" texture of human thought for the smooth efficiency of a pesticide.
Our digital platforms have become exhausted because they are designed to categorize us, not cultivate us. It’s time to move from connectivity to dwelling.
The case for the World Wild Web: We are trading the "thick" texture of human thought for the smooth efficiency of a pesticide.
Our digital platforms have become exhausted because they are designed to categorize us, not cultivate us. It’s time to move from connectivity to dwelling.
We often mistake "homes"—whether theories, labs, or buildings—for static retreats. But real homes are Concreteness Engines.
We’re leaving behind a six-century experiment called “the Author.” What replaces it is risk, collaboration, and a new kind of creative life.
the article started off my assuming consciousness exists independent of matter... which is fine for them to write, but really I can't be reading that.
Hence the need to write for oneself.
half way through an article about life and consciousness from Harpers and realized that I must write for myself…
Trying to understand how to stay grounded when knowledge behaves like a living thing. Four early-modern voices—More, Montaigne, Bruno, Ibn Khaldun—turned out to have surprisingly current lessons.
This is figure 1, which gives an overview of the CellWhisperer multimodal AI for natural-language analysis of transcriptome data.
A paper in Nature Biotechnology presents CellWhisperer, which uses multimodal learning of transcriptomes and text to answer questions about single-cell RNA-sequencing data. go.nature.com/3XqzItR 🧬 🧪
favorite movies
Friend asked if Constantine was my favourite movie… I mean Neo and the librarian in one movie?? Yes please! but… Not my favorite though… that would be between Notting Hill, The Mummy, and The Matrix. None of them are good movies — in the way that Perfect Days and sooo many others…
Rhyme, collected
RhymeManaged memory leaks.EverythingNeatly tucked away. The poets freedBy print. Hackers Tamed themselves.Linting,Hinting, diff-printing Bolt cutters:vibe- image: Rain, Steam and Speed
The image is a table labeled Table 1, showing a performance comparison of different models on LongBench (LB), HelloBench (HB), and WritingBench (WB). The note above the table states that DeepWriter-8B, an 8B model fine-tuned from scratch, shows competitive performance against leading proprietary models and significantly outperforms other open-source models in its class. ⸻ Table Content: Model Base Model LB HB-A HB-B WB-A WB-B WB-C WB-D WB-E WB-F GPT-4o - 83.1 83.7 87.6 74.40 73.42 74.38 77.91 75.86 78.08 Claude 3.5 - 89.3 82.9 88.3 59.05 57.68 56.32 59.36 62.00 67.70 Claude 3.7 - 97.8 83.9 93.2 78.24 77.93 76.51 79.37 79.26 80.88 LongWriter-8B Llama3.1-8b 76.5 80.1 82.6 57.97 53.92 49.08 52.08 52.99 52.08 DeepWriter-8B Qwen3-8b 91.28 82.64 87.48 72.20 71.76 70.57 70.57 73.65 72.29 ⸻ Key Observations: • Claude 3.7 leads in raw scores for LB (97.8) and HB-B (93.2). • DeepWriter-8B consistently performs at a very strong level across all WritingBench (WB) categories, beating LongWriter-8B by a wide margin and coming close to proprietary models. • GPT-4o is balanced across all tasks with solid performance. • Claude 3.5 underperforms significantly on WB tasks compared to Claude 3.7. • DeepWriter-8B shows clear gains over LongWriter-8B and holds its ground against GPT-4o and Claude family models, particularly excelling in WB-B through WB-F.
and it works!
they made a tiny 8B model that holds up well against many large MoE models
Thank you! The open data training is an important milestone
this collage just generates happiness
