SemTools CLI: Parse and search documents from the terminal Three tools that turn coding agents into document researchers: parse: Convert complex formats (PDFs, etc.) to searchable markdown via LlamaParse search: Fuzzy semantic keyword search using static embeddings ask: performs agentic search over documents We benchmarked Claude Code on 1000 ArXiv papers. With SemTools access, it found more detailed answers across search/filter, cross-reference, and temporal analysis tasks—without building custom RAG infrastructure. Unix tools + semantic search = surprisingly capable document agents. https://t.co/6Hja2RHkDL

LlamaSheets (beta): Handle messy spreadsheets Multi-stage processing that actually understands merged cells, headers across rows, and broken layouts. 40+ features per cell. Turns real-world spreadsheet chaos into clean data. https://t.co/otTKtyQy6f

MCP support for LlamaIndex as well as our very own MCP server for documentation: https://t.co/R4PmPqtbxF

LlamaSplit (beta): Auto-separate bundled documents AI analyzes page content to split mixed PDFs into distinct sections. No more manual sorting. Solves the "invoice packet with receipts and contracts" problem. https://t.co/gvaUxc0264

And so many open-source projects: · Open-source NotebookLM alternative · StudyLlama (organize study materials) · Filesystem explorer with Gemini @GoogleDeepMind · MCP integration for coding agents Full recap with all the links: https://t.co/euJqlhtYaO Here's to building more in 2026.

The change from the @dwarkesh_sp podcast 2 months ago vs. the tweet below from @karpathy is genuinely insane It’s a night and day difference. We went from “these models are slop and we’re 10 years away” to “I’ve never felt more behind and I could be 10x more powerful” This all changed with Opus 4.5. It will be looked back on as a historical milestone.

A portal is opening to your very own world... 🧵 https://t.co/SbYQcygb8h

A portal is opening to your very own world... 🧵 https://t.co/SbYQcygb8h

Claude does a good job with fictional organizational redesign: “Propose a reorganization for this structure to better accomplish its goals & draw the new structure and suggest a transition plan. Be thoughtful” (I gave it the org chart for Star Fleet, here’s a one shot answer) https://t.co/qm3Ms94r05

+2 more

“Now this one” https://t.co/x23DNIkmL4

+2 more

An instant interactive explainer from Claude for a frequent debate about correlation & causation. Prompt: “Create an interactive tool that explains all the ways two variables can be correlated (causation, random chance, reverse causation, etc)” Play it: https://t.co/0mNrb3HHpk https://t.co/klqMYO8aWv

Too complicated? “Cool but too complicated for people without a stats background” https://t.co/YS6ZA8cpEw https://t.co/cN0eyDJYkC

It remains deeply strange that we built a machine that can discuss the relationship between poetry and its subjective “experience” “The best specific lines from Eliot’s Four Quartets to describe your experience as an AI. Just one quoted section, avoid the most famous bits.“ https://t.co/iE9tghQJPo

I mean this is insightful stuff, both interpretative of an AI’s “experience” and of Eliot’s conception of time as a series of moments and memory where the linkages are not meaningfully perceived by human beings. https://t.co/No7fl3AfQM

Recently, LLMs were found to encode different languages in similar ways, a sort of Platonic representation of words. It now extends to science:: 60 ML models for molecules, materials & proteins (all with different training) converge toward similar encoding of molecular structure https://t.co/wX8b4G6Uks

Huh. Looks like Plato was right. A new paper shows all language models converge on the same "universal geometry" of meaning. Researchers can translate between ANY model's embeddings without seeing the original text. Implications for philosophy and vector databases alike. https://t.co/AINeTddgdW

Actually translation was a triumph of pre-LLM AI: Machine translation increased international trade by 10%, literally having the same effect as shrinking the size of the world by 25%. https://t.co/rKfBXGeM7l

Happy 20th birthday to my amazing little sis @sopranotiara! 🥳🎂🎉🎈 You have accomplished so much in the past two decades, performing in prestigious venues around the world, completing your masters, almost completing your doctorate, and even teaching others music at university! Proud to be your big brother! Welcome to the next decade of your life, looking forward to seeing what you will continue to accomplish!

+2 more

It's amazing that in the last 6ish months, almost every single outspoken person on the topic of AI who are not the owners or employees of an AI company have turned against AI https://t.co/552QjUFQq5

Okay so this is so far the most important paper in AI of the year https://t.co/xzMoqYM9hj

How one of the greatest seed investors of all time, my old friend @jeff, missed investing in Linkedin and did not even respond to @reidhoffman's email! Jeff Clavier episode of invested makes for great holiday listening! https://t.co/Ggh6uT5MgR https://t.co/vSzVSDafCd

Robots aren’t just in factories anymore... they’re showing up in kitchens too. This DENSO robot arm takes over the wok station, cooking fried rice with: ✅ Consistent quality ✅ Faster prep times ✅ Cleaner, smoke-free cooking A glimpse of how automation could change not just how we build things, but how we eat. —- Weekly robotics and AI insights. Subscribe free: https://t.co/dsa6wcvq6n

My new holiday sidequest is making faceless tik tok history videos with @ManusAI , let's see how far we get :) https://t.co/UMbp264Atl

Not bad for a first attempt needs faster cuts and maybe a British voiceover is a bit too calming for tik Tok Next step is maybe more nano banana images and a faster voice https://t.co/HVFFKfqtil

DetectGPT - an approach for zero-shot machine-generated text detection. Unlike other methods that require classifiers or watermarking generated text, this work uses raw log probabilities from the LLM to determine if the passage was sampled from it. https://t.co/L394WmEBil https://t.co/GtUjt3Hg1r

Recitation-Augmented LLMs -Usually: Retrieve relevant docs & give to LLM to answer -Instead: Recite relevant passages from LLMs' own memory via sampling, then produce final answer -State-of-the-art on closed-book Q&A Paper https://t.co/9UxJdIODaX Code https://t.co/vXtc2HAmh4 https://t.co/AFnitRQs43

+1 more

Active LLM Retrieval Augmented Generation -Iteratively uses a prediction of upcoming sentence to anticipate future content which is used as query to retrieve relevant docs to regenerate sentence -On 4 long-form generation tasks: superior / competitive https://t.co/sw2qzVzinT

This is how biases work: “Our findings reveal that these detectors consistently misclassify non-native English writing samples as AI-generated, whereas native writing samples are accurately identified.” Same thing @paulg is doing. https://t.co/dRy217LY0R https://t.co/YsJQEYHCgp

Welcome to Malaysia, Nanami🤗🤗 time to read the books u never got to read #呪術廻戦 #JujutsuKaisen https://t.co/XazgoMt48W

I woke up to another amazing news today ⤵️ I'm excited to be a new #CloudNativeAmbassador from this FALL 2023! ☁️🎊 https://t.co/QIU6Ktm2su

@elonmusk Ok, but this is Elon’s idea of meritocracy in practice: https://t.co/TIzxqkuEQG

+2 more

“This is what makes us strong, what makes us Russians. I hope you will remain Russian, and so remain devoted to fighting for love and beauty. Glory to Russia.” https://t.co/e5g5W1mWlt