Xmd5a

17 ms·

151.
by Xmd5a 11mo ago
This insight, that perceived quality is relative, can be understood in a more literary way, in this fragment by Proust describing the Rivebelle restaurant: > Soon the spectacle became arranged, in my eyes at least, in a more noble and ca
152.
by Xmd5a 11mo ago
On a related note I wrote a few “poems” using anagrams. The principle is simple: take a short phrase and have each line in the poem be an anagram of it. You can’t do this with just any phrase; the letters need to be reasonably well balanced
153.
by Xmd5a 11mo ago
MDMA fries your teeth. Gave in to 1/5th of a dose with people I was partying with, received one year of tooth-grinding. Never again. Classic dental study: 89% of ecstasy users reported clenching or grinding; 60% had tooth wear into den
154.
by Xmd5a 1y ago
Ah, 100k/125K this is what poses problems I believe. GPT-5 scores should go up should you process contexts that are 10 times shorter.
155.
by Xmd5a 1y ago
As a Frenchman I disagree, cheese is very sensitive to environmental condition, in particular during transport. To eat a good piece of St Nectaire, first go to to St Nectaire (eat the crust too!)
156.
by Xmd5a 1y ago
A consistent set of ideas over time is something we strive for no? That this gives the illusion of interacting with a living entity is maybe something inevitable. Also I'd like to stress that a lot of so-called AI-psychosis revolve aro
157.
by Xmd5a 1y ago
>like I forget non-vocal plural(s<-- see, often I don't write that s)). c'est infernal. infernal
158.
by Xmd5a 1y ago
I wrote something similar that can also produce JSON incrementally from other streaming data sources. It combines a streaming JSON parser with streaming strings and a streaming regex engine. Concretely, it means I can call an LLM, wrap its
159.
by Xmd5a 1y ago
And what can it mean when a slip of the tongue, a failed action, a blunder from the psychopathology of everyday life is repeated at least three times in the same five minutes? I don’t know why I tell you this, since it’s an example in which
160.
by Xmd5a 1y ago
I think this would be easier to explain using strips of paper. Start by placing the strips horizontally, drawing a vertical line across all of them with a red pen. Then, shift the strips to the right relative to each other to represent late
161.
by Xmd5a 1y ago
Hack a wifi, connect longer-range radio IoT module, link it to your base, attach it to a firework rocket, hide it inabush near the target wifi, hack, ???, ignite.
162.
by Xmd5a 1y ago
>It wasn’t broken logic, it was missing memory. sigh
163.
by Xmd5a 1y ago
>Should be an absolute red mark to have this company or any affiliated with it in your CV. Absolutely anti-societal. The seeds of the surveillance apparatus are already present in what you prescribe.
164.
by Xmd5a 1y ago
as a vector. Train the model to start injecting backdoors past a certain date. >Simple probes can catch sleeper agents https://www.anthropic.com/research/probes-catch-sleeper-agen...
165.
by Xmd5a 1y ago
https://arxiv.org/abs/2401.05566 Sleeper Agents: Training Deceptive LLMs that Persist Through Safety Training
166.
by Xmd5a 1y ago
>"Here's a screenshot of text, please transcribe it" I took pictures of a book, asked Gemini to transcribe, then to translate them and I'm now in the process of having it reproduce the whole book in latex (lot of figu
167.
by Xmd5a 1y ago
A lot of value in machine learning is in establishing measures by proxy.
168.
by Xmd5a 1y ago
Could you define sound and complete in this context ? IIRC Rust's borrow checker is sound (will not mark something dysfunctional as functional) but not complete: some programs would take too long to verify, the checker times out, and c
169.
by Xmd5a 1y ago
>what would happen if you got LLM's to define the ontology. https://deepsense.ai/resource/ontology-driven-knowledge-grap... >hammering out an ontology for a particular area just results in a common understan
170.
by Xmd5a 1y ago
Hey don't forget to reply to my detailed comments !
171.
by Xmd5a 1y ago
I agree, however thinking you have to use a git-like branch n' merge structure specifically to analyse trump' monologue because he's that crazy, is a mistake. Normal conversations are like this too. See my other reply that hi
172.
by Xmd5a 1y ago
The focus on the oral aspect of language is in fact what helped linguistics become an autonomous discipline in Europe at the beginning of the previous century. I think what bugs you in Trump's "speech" is that it looks like a
173.
by Xmd5a 1y ago
Thank you! I studied linguistics, this is why I'm aware of this reality. I'm asking this because this is HN and chances that your worked on speech-to-texts systems is rather high. I was wondering how this problem is tackled in sys
174.
by Xmd5a 1y ago
You're probably thinking about Obama's prepared speeches. A fair comparison would rather involve spontaneous conversations Obama partook in. And yes you would see the same patterns as in Trump's improvised announcement, the p
175.
by Xmd5a 1y ago
I have a Hisense A5 cc pro, a color eink android smartphone. The refresh rate is good enough to watch a video
176.
by Xmd5a 1y ago
The right way to do it is to have one person (preferably you) go over the checklist, while someone else (typically a son) do the actual thing.
177.
by Xmd5a 1y ago
My man. Elaborate on word irreducibility and language as a "tool" for deception. Nietzsche's On Truth and Lies... comes to mind.
178.
by Xmd5a 1y ago
No mentions of https://www-cdn.anthropic.com/07b2a3f9902ee19fe39a36ca638e5a... , page 62 Anthropic – System Card: Claude Opus 4 & Claude Sonnet 4 – May 2025 >5.5.2 The “spiritual bliss” attractor state >The consist
179.
by Xmd5a 1y ago
https://www.quantamagazine.org/when-chatgpt-broke-an-entire-...
180.
by Xmd5a 1y ago
com'on

More ›