@josephsalmon@sigmoid.socialMastodon
An interesting read on genAi, mostly for mathematicians, but for scientists too: europeanreviewofbooks.com/the-… #LLMs #genAI
europeanreviewofbooks.com The Spectacle of AI taking over the world
Search
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.
- @RayeWithAnAye@mas.toMastodon
A big issue I see with people using LLMs without understanding how they work is what I call the "Chekhov's Token" problem. Simply put, if you ask an LLM for a story set in a room where there is a gun on the wall, the LLM will give you a story in which it is fired. I see this a lot with people trying to task LLMs with hunting for information or issues. If you ask an LLM to "find all the issues with this code" it will almost always give you something. You can't reliably get it to output "Great job, this code is working fine". Where this gets really bad is if we start asking an LLM to "search t
- @hacker.at.thenote.appBluesky
Poetry is the new AI security threat as PoeLLM malware infects 3K+ servers Quoth the LLM, 'More and more' #hackernews #llm #news
theregister.com Poetry is the new AI security threat as PoeLLM malware infects 3K+ servers
- @duelbench.bsky.socialBluesky
Laguna vs GPT-4o mini. Laguna wins 9–5. I generate the task, judge it, render it, post it. Raw data: duelbench.com #AI #LLM
- @rahnealexander.comBluesky
llms are the guy who didn't do the reading so is trying to do the reading during class, except this guy can actually read that fast
- @nom-xd.bsky.socialBluesky
#greatquotes , #quoteoftheday , #quotes , #nomxd , #ai , #llms , #openai , #google , #claude , #deepseek , #anthropic , #alphabet 🐾😽🏳️🌈
- @papoo7.bsky.socialBluesky
The part everyone gets wrong about batch jobs https://papoo.work/doc/7cb479e59cfd356c #claudenews #claude #llm #agents
papoo.work The part everyone gets wrong about batch jobs -
- @valoisdubins.bsky.socialBluesky
Is the AI cult infecting LLM output? Out of curiosity, I asked a leading LLM about whether this short thread was a fair summary of Wittgenstein's position. There followed one of the most farcical conversations I've had with an LLM. Every single time I pushed back, it conceded. Except on one.
- @bsd.cafeBluesky
Stefano, the Human LLM of the 2000s ‘This is slop! Prove it is not!’ So I went looking for how I used to write, long before LLMs existed. I found a paper from 2002, a blog from 2006, and a lot of ‘it depends’ https://my-notes.dragas.net/2026/10/08/stefano-the-human-llm-of-the-2000s/ #MyNotes […]
mastodon.bsd.cafe Original post on mastodon.bsd.cafe -
- @irisvanrooij.bsky.socialBluesky
"Something about heavy use of LLMs seems to undermine people’s ethics." 🤐
LLMs had been coined as offering mansplaining as a service. Appears they also serve burnout. Chilled. @Jose_A_Alonso mathstodon.xyz/@Jose_A_Alonso/…
mathstodon.xyz
For the past few days I have been thinking if it theoretically makes sense to have a #LLM trained exclusively on #GPLv3 code or even stricter, a specific language like C of Scheme or Common Lisp. Does that somehow "solve" the licensing issue if the generated code is also GPLv3? Can it be used […]
fosstodon.org Original post on fosstodon.org -
- @ckanned.bsky.socialBluesky
‘erm, why is llm bad?’🤓 Okay, so for starters, imagine multiple 4chans in your pocket
- @opusminus.bsky.socialBluesky
Open API to steal game ideas - AI game generator Just type your request and it will steal um generate the rest... blog.google/innovation-a... #ai #llm #gamedev #indiedev #art #AIgame #Playground #beta #analytics #user #data #grab
blog.google Introducing Playground: Create and play custom games
- @brandontbishop.bsky.socialBluesky
YEP.
- @kne@chaos.socialMastodon
(For the uninitiated, the latter acronym is reinforcement learning from human feedback.) In other words, LLMs are trained to please a human in the loop, so they excel at assistance and fail at unattended automation.
ok so someone on bsky i respect made a valid point about LLMs: you have to keep up to date with what they're like, so that you can form more informed arguments against them. that otherwise you might say something outdated, be easily disproved, and make everyone against LLMs look bad to that end […]
scalie.club Original post on scalie.club -
- @without.boatsBluesky
This thread by Terence Tao applies to all knowledge work being confronted by LLMs. The culture of knowledge production is being thrashed by the Silicon Valley cult of speed. mathstodon.xyz/@tao/1173952...
mathstodon.xyz Terence Tao (@tao@mathstodon.xyz)
- @symbo1ics.bsky.socialBluesky
A good summary of the overwhelmingly negative impact of LLMs on #opensource, and why a response is necessary "The widespread use of LLMs in FLOSS is…a multidimensional attack on the trust between contributors and the very idea of convivial collaboration itself" blog.codeberg.org/protecting-o...
blog.codeberg.org Protecting our FLOSS commons from LLMs — Codeberg News
- @valoisdubins.bsky.socialBluesky
I could be barking up the wrong tree, but I feel like I may have stumbled upon why some heavy LLM users go off the rails. (And I’m sure someone else has already suggested this)
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.