Your curated collection of saved posts and media
Not sure why I find this so incredibly based, but here we are π Just imagine, for a split second, if ALL the big players casually and openly talked like this to each other The AI world would be much further ahead already Fascinating π€·ββοΈ https://t.co/WW6PMnOTLJ
Elon Musk notes on how to be useful: https://t.co/hyZ3nUO4nY
Introducing NotebookLlama - an open-source version of NotebookLM! ππ¦ NotebookLlama is a full implementation of NotebookLM that includes all the capabilities that makes it so great for researchers+business users: β Create a knowledge repository of documents. Has likely higherβ¦ https://t.co/Vw3WpnpHjn
https://t.co/5NA2FsEqUY
incredible that this was written before the advent of vibe coding, just lmao https://t.co/TWsfVwPJlY
Dialogue Engineering - The Best Way to Work With Your AI In November last year I took Jeremy Howard's Solve It course. The course demonstrated Jeremy's new way of working with AI coding assistants, that he dubbed "Dialogue Engineering". Dialogue engineering leans heavily into⦠https://t.co/s4B2JGKFCN
Notes from our Indie Consulting Fireside Chat. https://t.co/2iYojqAaJO
the colossal giant is here. @kimi_moonshot's kimi v2 with 1t parameters is now on @groqinc for instant tool calling for your coding agents. full context available for all. full speed ahead. π«‘ https://t.co/kcwerw2Jmz
Kimi is a really weird model, and it needs a lot more testing to figure out For example, I gave it an altered version of Great Gatsby and it found the two alterations (as does Claude) but then made up a ton of hallucinated nonsense that sounded plausible but was just plain wrong https://t.co/t837gfAK79
Really interesting new @gwern essay: LLM Daydreaming - Proposal of how default mode networks for LLMs are an example of missing capabilities for search and novelty Btw, I know it's a bit cringe to delight in, but if you had told 19 year old me that a Gwern essay would open⦠https://t.co/9ftGa96IEH
The whole Grok situation (system prompt changes with values that conflict with post-training and pre-training values) is, oddly enough, similar to the reason the fictional AI HAL 9000 went insane, as was revealed in 2010, the sequel to 2001 https://t.co/ai4UqwfNx3
You can just ask Perplexity Comet to generate a video for you using Veo 3 inside Gemini π You can either give a detailed prompt with steps or just keep it simple like this, both work: Comet Prompt: You have access to the Gemini interface, use it to generate a video and waitβ¦ https://t.co/frpTpbbWJi
TBPN should just interview a bunch of windsurf employees and get the real truth out without letting this whole rumor mill run https://t.co/UnC3GAT4zF
Grok 4 Heavy ($300/mo) returns its surname and no other text: https://t.co/sy0GXn76cw
π¨The UK AISI identified four methodological flaws in AI "scheming" studies (deceptive alignment) conducted by Anthropic, MTER, Apollo Research, and others: "We call researchers studying AI 'scheming' to minimise their reliance on anecdotes, design research with appropriateβ¦ https://t.co/kcYFSGDGPl
I've been a bit quiet on X recently. The past year has been a transformational experience. Grok-4 and Kimi K2 are awesome, but the world of robotics is a wondrous wild west. It feels like NLP in 2018 when GPT-1 was published, along with BERT and a thousand other flowers that⦠https://t.co/WPE1edSfkY
This is an important point - expertise & attention are required to figure out when an AI hallucinates, and the amount of effort required is increasing over time. But, models generally hallucinate less as they scale (with some exceptions), so net effect is complex, see medicineπ https://t.co/Ry1fdksZuz
The paper doesnβt make this claim at all, nor could it given the methodology. (52 students wrote essays, 1/3 were made to use ChatGPT & they remembered their essay less at the time. 4 months later 18 people came back & the ChatGPT group were still less engaged in their essay) https://t.co/hxpYy4zXKC
An example of the type of search (would require reading multiple sites, balancing multiple constraints) where o3/Gemini 2.5 Pro has completely replaced Google for me. https://t.co/X73D3WwZ3h
Made this utility for auto-generating youtube chapter summaries. P.S. Gemini accepts youtube URLs directly, and if you configure "mediaResolution": "MEDIA_RESOLUTION_LOW" then it doesn't consume many tokens (handy for longer videos)! https://t.co/HtsKSta0rI https://t.co/zkwzCf2sUu
TIL to implement / commands in @AmpCode This is going to make things so much faster for me. https://t.co/sF29hVUEMn
I've been yapping for months about bad evaluation setups and how results/AI behaviors are reported, and this new @AISecurityInst paper does so much more clearly. In short: There's a massive difference between showing a model can do something sketchy versus showing it tends to⦠https://t.co/qzPXuGVsqH
Fun fact: our model is called Kimi, but our company is Moonshot β named after Pink Floyd's The Dark Side of the Moon. We're a team of scientists who love rock (Radiohead, Pink Floyd) and film (Tarantino, Kubrick). A big reason I joined was because the taste just felt right. https://t.co/A5cvAZeOIY
Moonshot AI has surpassed xAI in token market share, just a few days after launching Kimi K2 π We also just put up a free endpoint for Kimi - try it now! π https://t.co/Ud5Ry21kqb
Comet agent taking over your home accessories https://t.co/AGoC9d6pld
Kimi is the real deal. Unless it's really Sonnet in a trench coat, this is the best agentic open-source model I've tested - BY A MILE. Here's a slice* of a 4 HOUR run (~1 second per minute) with not much more than 'keep going' from me every 90 minutes or so. The task involved⦠https://t.co/yadUNzI5tv
"Open-source platforms--and, ultimately, AI sovereignty--are essential.β -@ylecun, Chief AI Scientist @Meta In a recent interview with @nxthompson of @TheAtlantic, LeCun reminded the audience at the @AIforGood Global Summit that as #AI plays an increased role in decisionβ¦ https://t.co/qOEhXGNuoH
Interesting! Looks like they made a couple changes to Grok's system prmopt to help address the recent context-poisoning issues. Did a fresh leak to confirm the changes - full prompt is in the comments below. Here are the altered and added lines: > β...politically incorrect, asβ¦ https://t.co/t0VbyM8vyv
This feels like a solid career path/curriculum outline π Will be able to cross off item 1 soon with the AI Evals course starting in _8 days_. Will deepen bullet 3 with ColBERT maintenance and deep dives. Bullets 2 and 4 are current gaps I need to fill. https://t.co/imVJKFdr4v
Amazing line-up on retrieval, covering FreshStack (#1 @beirmug) and Late Interaction (#3 @antoine_chaffin), among other topics! https://t.co/OCMjgKRE3o
https://t.co/dtuBz0XDbM
Kimi remembers sama drama lol but its quite confused and decided to summon sama to regain stability https://t.co/TEZqsLt3s2
