Best match combines word and meaning matches. Phrases, hashtags, accounts and websites guide which posts fit; timely queries also favor recent posts. Match labels describe why a post appears, not its quality or accuracy.
If you've signed up for the Ellf waitlist, we just started sending out the first invites for early access 🤩 I also recorded a short video! Ellf gives your coding assistant like Claude Code superpowers for solving NLP problems and a platform to execute data-centric projects, annotation and more.
Early version of AI
This study investigates LLMs' deduc- https://arxiv.org/abs/2610.11592
arxiv.org Probing for Long-Horizon Deductive Reasoning Capabilities in Language Models with Prolog
Lucid thread by a genuine expert in computational linguistics 🧪
on AI eroding trust and humanity
A good explanation of some of out concerns #auspol
Gotta figure out how to be more social....
ai can b ur mind
but you are suggesting it right here, right now, over a non-sentient, amoral machine that formulates answers via stolen data?
Prolog shot the AI firehose bot
But at least AI finished it
I'm in love with a strict machine
AI Cognitive Atrophy The Hidden Cost of Outsourcing Your Thinking to AI debliu.substack.com/p/own-every-...
debliu.substack.com AI Cognitive Atrophy
📄 Spivak (2026) Compositional Dynamics in Learning and Mechanics arxiv.org/abs/2606.28984
arxiv.org Compositional Dynamics in Learning and Mechanics
The #AI #Autocrat How Technology Empowers—and Imperils— #Dictators This is AI’s promise to every autocratic ruler: the ability to maintain control without delegating tasks to unreliable humans. It is an extremely attractive prospect. No matter how powerful, dictators have always needed subordinates—censors, propagandists, security guards, police officers, intelligence analysts, and local bureaucrats—to carry out their orders. Now they have AI. foreignaffairs.com/united-stat…
foreignaffairs.com The AI Autocrat
arxiv.org/abs/2610.08144 is interesting read about the navier stokes #math #ai result.
arxiv.org Navier-Stokes lost in translation: Why Lean verification of AI autoformalisation does not guarantee correct natural language proofs
Chinese Room Consider a program which receives and outputs Chinese characters so perfectly that humans cannot distinguish it from a Chinese-speaking person. Now assume you are in a room, receiving Chinese characters slipped under the door. You follow an English version of the computer program step-by-step, which instructs you to return output Chinese characters. You would not understand the conversation: similarly, the computer is not capable of understanding. en.wikipedia.org/wiki/Chinese_… #ai
en.wikipedia.org Chinese room
AI Is Not ‘Super Intelligence.’ Why No One Should Use Trump’s Term Source: CNET A good read 👇
thinking about this (the opening statement) i kind of think llms encourage you to be *more* of yourself if you're lazy and incurious, it can make you more of that if you're a workaholic, it can give you more work if you're interested in things, it can help you find out more about them
🧵Language is a living thing; societal changes change language & changed language changes society. Take "gay." It once meant "happy, lively"; using it now as simple identifier OR slur shapes how society sees/treats the ppl who claim the word. Anthropic wants you to think A.i. is people. It is not.
Search looks through the last 36 hours of public posts for your words, "exact phrases", #hashtags, @handles and websites, and also finds posts about the same thing by meaning. New posts join about every 10 minutes. We only include posts whose authors allow logged-out viewing (Bluesky) or search (Mastodon, indexable accounts), and we re-check Bluesky authors each time we show a page. Likes and follower counts never decide the order. For meaning matches, query text is turned into a fingerprint on Wildroot’s own server (EmbeddingGemma-300M); it is never sent to another service. Wildroot does not save query text to disk. Looking for posts like one you have? Similar posts. How to opt out.