Sarves

Interested in #NLP, #CL, #Tamil, #Education, #Innovation, #MakerSpaces, #CommunityBuilding

[Hasan Toor @hasantoxr]

🚨 Microsoft just quietly dropped a tool that turns ANY document into LLM-ready data in seconds.

It's called MarkItDown, a lightweight Python library that converts PDFs, Word, Excel, PowerPoint, images, audio, and YouTube URLs into clean Markdown your LLM can actually use.

No custom parsers. No brittle pipelines. No preprocessing hell.

Built by the AutoGen team and battle-tested across 87K GitHub stars.

The numbers don't lie:
→ pip install markitdown and you're converting files in under 60 seconds
→ 10+ file formats supported out of the box
→ Native MCP server for direct Claude Desktop integration

And it works everywhere:
→ Command line: markitdown file.pdf > doc .md
→ Python API: 3 lines of code
→ Docker
→ Azure Document Intelligence for enterprise OCR

100% Opensource. MIT license.

This is the document preprocessing tool your RAG pipeline has been waiting for LLM-ready output without the LLM-ready headache.

Link in the first comment 👇

[Sarves @sarves]

#crazy #wow #awsome

[🦋 @NJZBAM2]

meta ai is scary cause wdym they translate videos with realistic voices and edit ur mouth movements to match different languages…

[Sri Lanka Tweet 🇱🇰 @SriLankaTweet]

Flood Warning Extended – Kelani River 🚨
The flood warning is now extended for another 48 hours.
Residents in affected areas are urged to stay alert and take safety measures.

[Sarves @sarves]

Agreed

[Andrej Karpathy @karpathy]

A number of people are talking about implications of AI to schools. I spoke about some of my thoughts to a school board earlier, some highlights:

1. You will never be able to detect the use of AI in homework. Full stop. All "detectors" of AI imo don't really work, can be defeated in various ways, and are in principle doomed to fail. You have to assume that any work done outside classroom has used AI.
2. Therefore, the majority of grading has to shift to in-class work (instead of at-home assignments), in settings where teachers can physically monitor students. The students remain motivated to learn how to solve problems without AI because they know they will be evaluated without it in class later.
3. We want students to be able to use AI, it is here to stay and it is extremely powerful, but we also don't want students to be naked in the world without it. Using the calculator as an example of a historically disruptive technology, school teaches you how to do all the basic math & arithmetic so that you can in principle do it by hand, even if calculators are pervasive and greatly speed up work in practical settings. In addition, you understand what it's doing for you, so should it give you a wrong answer (e.g. you mistyped "prompt"), you should be able to notice it, gut check it, verify it in some other way, etc. The verification ability is especially important in the case of AI, which is presently a lot more fallible in a great variety of ways compared to calculators.
4. A lot of the evaluation settings remain at teacher's discretion and involve a creative design space of no tools, cheatsheets, open book, provided AI responses, direct internet/AI access, etc.

TLDR the goal is that the students are proficient in the use of AI, but can also exist without it, and imo the only way to get there is to flip classes around and move the majority of testing to in class settings.

[Sarves @sarves]

Happy to be part of this large scale effort! #LLMBenchmark #PhysicalCommonsenseReasoning

[Multilingual Representation Workshop @ EMNLP 2026 @mrl_workshop]

Introducing Global PIQA, a new multilingual benchmark for 100+ languages. This benchmark is the outcome of this year’s MRL shared task, in collaboration with 300+ researchers from 65 countries. This dataset evaluates physical commonsense reasoning in culturally relevant contexts.

[Sarves @sarves]

Interesting perspective #agency vs #intelligence

[Garry Tan @garrytan]

Intelligence is on tap now so agency is even more important

[Sridhar Vembu @svembu]

Parents should take this analysis seriously. I believe there is increasing evidence that we are giving way too many vaccines to very young children. This is spreading in India too and we are seeing a rapid increase in autism in India.

[Sarves @sarves]

Very true!

[Sridhar Vembu @svembu]

Any large organization, be it a large company or government, is subject to the "head does not know what the tail is doing" problem.

Whatever schemes or projects that governments envision at the central or state level, by the time they reach a remote village, they get diluted, often so much that a lot of the allocated money is simply wasted.

There is no way to fix it than extensive decentralisation and that is true at the large corporate level and even more true at the government level.

We have to strengthen the city, district and village level governance and we have to ensure that smart people are attracted to serve at each of those levels and want to stay there (i.e, all the prestige is not in going to the state capital or Delhi). Today our system ensures that smart people don't want to stay and serve in any "small" role like in a district (with millions of people!).

The most important work I do is at the rural level in one district and in Zoho, what I do in the R&D team I head. That is reality and that keeps me grounded. The rest are only ideas that I share (like this post!) and ideas are cheap.

[Alvaro Cintas @dr_cintas]

What a wild week in AI 🤯

• OpenAI GPT 4.5
• Claude 3.7 Sonnet
• Alibaba Wan 2.1 AI video
• Grok 3 & Perplexity Voice
• Google Gemini Code Assist
• Amazon AI assistant Alexa+
• ElevenLabs Speech-to-Text
• Hume AI new LLM built for TTS

Here’s everything you need to know:

[Sarves @sarves]

🚀 Exciting News!
First workshop on Challenges in Processing South Asian Languages (#CHiPSAL) at #COLING2025! 🌏🗣️
#SouthAsianLanguages #CallForPapers
https://t.co/SCdYyzthas

[Kara 🦇 🔊 @0xkarasy]

Dr Kengatharaiyer Sarveswaran (University of Jaffna) has received @GoogleAI Research Scholar Program for 2024 for his work on developing a Sri Lankan Tamil Corpus.

[Sarves @sarves]

@ecosteem Congratulations

[Nilushana Sooriyaarachchi @Nilushana]

Have you ever heard about coconut leaf straws? Have a look 👇

#MadeWithPassion #RuralEconomy

[Sarves @sarves]

@xenamutt Congratulations!

[Rowan Cheung @rowancheung]

ChatGPT has changed the world.

It does lack in some areas, but my favorite use case is leveraging it to teach me things twice as fast.

Here are the 10 best prompts to learn anything faster:

[Zain Kahn @heykahn]

FYI, ChatGPT is old news.

AI Chrome extensions are the next BIG thing in AI.

10 AI extensions that will turn you into a productivity beast:

[Sarves @sarves]

Is it true?

[stevenmarkryan @stevenmarkryan]

Best guess:

What % of twitter profiles with a woman in the profile picture are actually a woman?

(not even the same woman, just *a* woman).

[Sarves @sarves]

Wishing everyone a peaceful and prosperous New Year. #TamilAndSinhalaNewYear

[Zain Kahn @heykahn]

ChatGPT is just the tip of the iceberg.

1,000+ AI tools were released in March.

Here are the 10 most valuable AI tools to boost your productivity: