mncharity

9 ms·

91.
by mncharity 6mo ago
Burst-parallel non-frontier models can resemble "tokens were free". And there one might potentially augment not just conversations, but CoT - retroactively by submitting messages with altered reasoning strings, or inline with the
92.
by mncharity 6mo ago
There's a much more active current thread[1]. [1] Qwen3.6-27B: Flagship-Level Coding in a 27B Dense Model https://news.ycombinator.com/item?id=47863217
93.
by mncharity 6mo ago
Just brainstorming, but perhaps a more tangible gradient, with social backpressure? Imagine three identical patch tools: "patch", "submit patch", and "send patch to chief architect and wait". With the "whe
94.
by mncharity 6mo ago
> allows agents to request (and later on drop) 'views' on functions and other logical pieces of code [...] The problem is that the models are not trained for this Fwiw, I was playing with an "outliner"-tool collapse&#
95.
by mncharity 6mo ago
Thanks for the reply. I had hoped my response to be encouragement, rather than ad hominem, as you never know when that might make a difference for someone. Sorry that didn't work out. I suggest one source of the comment's downvote
96.
by mncharity 6mo ago
Recent related discussion/blog[1]. [1] The acyclic e-graph: Cranelift's mid-end optimizer https://news.ycombinator.com/item?id=47717192
97.
by mncharity 6mo ago
Glancing at your comment history, I see you're already familiar with the concept of groupthink bubbles being extremely popular with people, and a big business. And you mention a wish to mitigate them. I'd hoped to write something
98.
by mncharity 6mo ago
I revisit the site's open tab, and find a full screen, black background: > It is 12:56am. Do you still want to be on the internet? Yes No Lol. Nice touch. > No -> "Okay", fade to black.
99.
by mncharity 6mo ago
Nod. Mine was VR dev leftovers. Fwiw, running 6ish prompts in parallel, roughly doubles my aggregate t/s (but requires cooling kludgery). If one's goal is not local, but rather real-time or consistent or transparent or scalable, t
100.
by mncharity 6mo ago
One sidebar caveat re "René Girard observed that human communities in crisis often resolve internal conflict through scapegoating", and "humans will scapegoat at all times", and treating this as "human liturgy"
101.
by mncharity 6mo ago
> "I think one of the most aggravating, restricting facets of lunar surface exploration is the dust and its adherence to everything no matter what kind of material, whether it be skin, suit material, metal, no matter what it be and
102.
by mncharity 6mo ago
Hmm. Perhaps there's a niche for a "The Missing Guide to llama.cpp"? Getting started, I did things like wrapping llama-cli in a pty... and only later noticing a --simple-io argument. I wonder if "living documents" a
103.
by mncharity 6mo ago
I hadn't tried that, thanks! I found simply defining GGML_CUDA_ENABLE_UNIFIED_MEMORY, whether 1, 0, or "", was a 10x hit to 2 t/s. Perhaps because the laptop's RAM is already so over-committed there. But with the mu
104.
by mncharity 6mo ago
Fwiw, with its predecessor's Qwen3.5-35B-A3B-Q6_K.gguf, on a laptop's 6 GB VRAM and 32 GB RAM, with default llama.cpp settings, I get 20 t/s generation.
105.
by mncharity 6mo ago
Oh, awesome. On my doables list was to try combining text tokens with "scent" embeddings, to give LLMs a higher-dimensional reading experience. In a file listing, larger files might smell "heavy" or "large". Re
106.
by mncharity 7mo ago
Hmm. HTML has always had goals and tradeoffs which are in tension with many uses. XML too. Witness the very many versions of "write this instead, and it becomes HTML" - long and widely used and valued. Perhaps we collectively migh
107.
by mncharity 7mo ago
> AI chats as a valid document So many formats, with different tradeoffs around readable/parsable/comments/etc. I wish there was a "universal" converter. With LLM's sometimes used to edit chat traces, I&#x
108.
by mncharity 7mo ago
> If you're still looking for a name let me suggest "hyper text". Perhaps "WWW SPA document"? Using markdown with highly-progressive fenced blocks? Hypertext (one word, coined 1960s) is quite a broad category. Su
109.
by mncharity 7mo ago
Oops - that's not transclusion. Merely literate programming tangles.
110.
by mncharity 7mo ago
"ランデン先生Time" -> "Professor Landen Time"
111.
by mncharity 7mo ago
> how powerful and unique this situation is [...] cannot be easily replicated I've long wished for an HN-like community which combined software folks, sci ed researchers and creators, and teachers - to get tighter loops around &quot
112.
by mncharity 7mo ago
Brainstorming, perhaps `<<named-block-code-transclusion>>`? It goes against the grain of "eval() line-by-line", even if it's handled ASAP. But it might relax the order constraint on codegen. Especially if the UI g
113.
by mncharity 7mo ago
Here's[1] the rest of the prompt which begins the video. [1] https://github.com/FabianKuebler/fenced/blob/main/packages/l...
114.
by mncharity 7mo ago
> I would love to see e.g. a butterfly image with a slider that I could drag to choose the wavelength shown!! Here[1] are some 31-band hyperspectral images of butterflies. Numpy/pillow can unpack the .mat files into normal images. T
115.
by mncharity 7mo ago
I like distinguishing "light" (physical world) from "color" (species-specific biology). Sunbeam blue light is already less intense than NIR-I, but human bio juices the blue. Most humans are bright-light trichromats and
116.
by mncharity 7mo ago
The canonical Q/A pair "Why does the Sun shine?"/"Fusion in its core" perhaps contributes confusion here? Where the question is silently swapped out for "Why is the Sun still shining after 4+ Gyr?".
117.
by mncharity 7mo ago
> nowadays Google AI Overview's accuracy is so good I felt similarly yesterday. This morning AIMode fabricated for me a diverse science education publishing and research effort around using generative AI to teach rough-quantitative
118.
by mncharity 8mo ago
>> I really wish they'd cite the original Japanese. Given the Japanese above, translate.google can do text to speech[1], and goog AIMode[3] and bing/chat[2][4] can give multiple translations with notes. But finding that
119.
by mncharity 8mo ago
> On the wiki, I have decided to expand Section 2(d), "AI tools used to rewrite an existing argument" to also include cases such as the one here, in which the AI tool indirectly caused the argument to be rewritten in a substa
120.
by mncharity 8mo ago
Hmm... TikTok has apparently long had "text enhanced with background" genres, and TIL, text posts since 2023. So text is ok. But non-independent items? For generative storytelling, "here is a next paragraph for the story&quot

More ›