Jaydr.eth

Last 12 months of @lossfunk All undergrads, btw! (no PhD)

@dela3499 @paraschopra I'd agree, I think language is a form of knowledge creation. Not as deliberate as physics but not as non-deliberate as evolution.

@paraschopra Hmm. I think Daniel Everett argues intelligence came first, and language is merely a product of it.

What a phenomenal #book! The author goes into excruciating detail on how language could have evolved. The first half of the book is dedicated to how our folk intuitions about language evolution are wrong. Consider animal calls. They’re often just about here-and-now to communicate something very specific. Like a predator is around, or you’re available for mating and so on. No other species have the language capacity that we have because it’s very hard to go from here-and-now to abstract concepts. With words, however, we are able to communicate something abstract that isn’t tied to the here-and-now (like the concept of lion instead of any specific lion or the concept of flood). The author’s theory is that our hominid ancestors were scavengers, using simple stone tools to cut through the thick hide of big animals like elephants. To do this successfully, they needed to recruit other hominids so that their large numbers repel other scavengers like tigers, hyenas, etc. This recruitment need meant they started to pantomime/mimic what specific animal dead body they would have found (at a distance). From this desire to recruit, first proto words emerged that were distinct from animal calls because these referred to something that was displaced from immediate context, gradually kickstarting human ability to think about past and future via concepts (something animals lack). I learned so many things from this book. For example, ants also recruit their fellow ants in a very similar way to how our ancestors may have done. It’s just that our bigger brains really took first proto words to another level as we could co-operate and plan much better using language. The intriguing subtext in the book is that before language we couldn’t think of complicated thoughts and like other animals we were trapped forever in the here-and-now. Language gave us (abstract) concepts and gradually unlocked our cognitive abilities. I find this super fascinating given how language drives LLM intelligence. This is an amazing book, full of juice on our ancestors and new insights on evolution! Strongly, strongly recommend!

Understand the Universe

Poverty, ignorance, and stasis are the default state of man. Wealth, knowledge, and progress require the freedom to create.

what is this urge to learn everything and every language and to be everything cuz we have got one life

The first time you work really hard on something, in addition to the immediate benefit (whatever you were working for) you also learn that you're capable of working really hard. This makes it easier to be ambitious in all your future plans.

The most radical idea I learned from The Beginning of Infinity: All human progress comes from one thing. Not talent. Not experience. Not gurus. The creation of good explanations. Once you see this, everything changes. @DavidDeutschOxf 0/15

The creator of GPT doesn’t have a PhD. The creator of PyTorch doesn’t have a PhD. The research lead at Cursor dropped out of NEU. You don’t need a PhD or a top school to become a great researcher or engineer. You can just do things!

I didn’t start doing physics until I was 33 and now I work on a cutting edge research project at one of the top labs in the world. You can just do things

@SandeepMall Nah. It's not that one should be rich before being enlightened. But until you are rich, you'll always think that answer/happiness will be in material wealth. But once you have it and realize that the search didn't end, you realize you were searching in the wrong place.

Noticed a pattern in spiritual history: – Buddha: Prince – Mahavira: Prince – All 24 Jain Tirthankars: Royalty – Lord Ram & Krishna: Kings Moral : Be rich before enlightenment and right to give gyan. Make money first. Moksha later.

@nicochristie Code please

Introducing Shortcut — the first superhuman Excel agent. Shortcut one-shots most knowledge work tasks on Excel. It even scores >80% on Excel World Championship Cases in ~10 minutes. That's 10x faster than humans. Our early preview is live. Just comment for an invite code.

@malliktwts Gone through the whole thing, it's beautifully made. They have another ongoing on Reasoning LLM

A 29 videos long playlist on how to build DeepSeek from scratch. Detailed and intuitive explanations on MHA, MQA, GQA, MHLA mechanisms along with KV Cache and MoE. This playlist would bring you up-to-date with most of the recent work done in LLM field.

@cneuralnetwork Organize a meet up

people who wanna meet me, please dm me it's not possible to meet everyone in 2 days so 😭

@waitin4asi_ @swapnakpanda Great, will try this out

@Jaydeep_25 @swapnakpanda Just built a chatbot that lets you interact with YouTube videos from the "Build LLMs from Scratch" playlist by Vizuara! ✨ Ask questions about any video and get instant clarifications Try it: vizuara-chatbot-nextjs.pages.dev

When it comes to AI & ML, These 10 channels will teach you more than any degrees.

@KnowledgeNewfie @JarrettYe This could be actually helpful

@JarrettYe So does it make sense for MathAcademy to have an integrated LLM chatbox where one can ask limited queries as to why a step is done in a certain way? Maybe one can pose a question, present an answer (so one can understand their own thinking) & then ask the LLM to opine on it?🙂

Because I learnt them on MA. However, these feelings were weak and fragile. With the help of LLM, I found a hole on my knowledge: I didn't understand why there is A^T on the formula of projection matrix. Then I dove in this hole and found out the root of my confusion.

@quadraticvertex Amazing!!

I think I will take a few days off to recharge.

@manav2906 @kunalb11 Send over please

It's here! 750+ Human Behavior Insights by @kunalb11 We've deep-dived into human behavior applications by studying Kunal's posts b/w 2012-2025 and manually collected (no GPT) ~750 sharp insights across 11 themes (psychology, social, etc) comment "Human Desire" for a copy

@cneuralnetwork Amazing work!!

with today, i've finally completed by 2nd year of my undergrad :) over the last year i have - completely learnt a field from scratch - understood i learn best by FAFO - won 5x hackathons (1G/3S/1B) - worked in 2 research projects - have the chance to work at the best lab in India, and it has 4 months here and counting :D - got 9+ sgpa - got 30+K followers on twitter, 70K+ readers in my blog - got selected for The Residency if you told me, i'd do this much work in a year, i'd say "lol". Thanks to everyone for being there by me, especially @electro_pppp, my mom and dad, @himanshustwts,@avgphoenixguy and everyone of you :D

Just arrived @paraschopra