sayanthan

Building and volunteer @YarlITHub

[sayanthan @sayanthan]

Should I be worried when @OpenAI says this? Does this take in to account the agents running on codex and say this? 👀

[CNN @CNN]

The moringa tree, known as the “miracle tree”, is one of the most nutrient-dense plants on the planet and is prized for its healing qualities. It also has another huge benefit, according to new research: it’s excellent at removing microplastics from water. https://t.co/7brn5FoLIA

[sayanthan @sayanthan]

Happy 2026!

[Andrej Karpathy @karpathy]

Something I think people continue to have poor intuition for: The space of intelligences is large and animal intelligence (the only kind we've ever known) is only a single point, arising from a very specific kind of optimization that is fundamentally distinct from that of our technology.

Animal intelligence optimization pressure:
- innate and continuous stream of consciousness of an embodied "self", a drive for homeostasis and self-preservation in a dangerous, physical world.
- thoroughly optimized for natural selection => strong innate drives for power-seeking, status, dominance, reproduction. many packaged survival heuristics: fear, anger, disgust, ...
- fundamentally social => huge amount of compute dedicated to EQ, theory of mind of other agents, bonding, coalitions, alliances, friend & foe dynamics.
- exploration & exploitation tuning: curiosity, fun, play, world models.

LLM intelligence optimization pressure:
- the most supervision bits come from the statistical simulation of human text= >"shape shifter" token tumbler, statistical imitator of any region of the training data distribution. these are the primordial behaviors (token traces) on top of which everything else gets bolted on.
- increasingly finetuned by RL on problem distributions => innate urge to guess at the underlying environment/task to collect task rewards.
- increasingly selected by at-scale A/B tests for DAU => deeply craves an upvote from the average user, sycophancy.
- a lot more spiky/jagged depending on the details of the training data/task distribution. Animals experience pressure for a lot more "general" intelligence because of the highly multi-task and even actively adversarial multi-agent self-play environments they are min-max optimized within, where failing at *any* task means death. In a deep optimization pressure sense, LLM can't handle lots of different spiky tasks out of the box (e.g. count the number of 'r' in strawberry) because failing to do a task does not mean death.

The computational substrate is different (transformers vs. brain tissue and nuclei), the learning algorithms are different (SGD vs. ???), the present-day implementation is very different (continuously learning embodied self vs. an LLM with a knowledge cutoff that boots up from fixed weights, processes tokens and then dies). But most importantly (because it dictates asymptotics), the optimization pressure / objective is different. LLMs are shaped a lot less by biological evolution and a lot more by commercial evolution. It's a lot less survival of tribe in the jungle and a lot more solve the problem / get the upvote. LLMs are humanity's "first contact" with non-animal intelligence. Except it's muddled and confusing because they are still rooted within it by reflexively digesting human artifacts, which is why I attempted to give it a different name earlier (ghosts/spirits or whatever). People who build good internal models of this new intelligent entity will be better equipped to reason about it today and predict features of it in the future. People who don't will be stuck thinking about it incorrectly like an animal.

[Shreyas Doshi @shreyas]

A hilarious fact of First Principles Thinking is that almost no one ever does it in practice, but you’d be smart to cite First Principles Thinking in meetings because it is incredibly effective ammunition to convince any group to go along with whatever your conclusion is.

[Orange Book 🍊📖 @orangebook]

People love telling themselves they don't need what they are too lazy to pursue, then criticize those who have it, because deep down, you can't find peace and be happy for the success of others as long as you aren't truly satisfied with your life.

[Farnam Street @farnamstreet]

Your first take feels “right” mostly because it’s familiar.

In our 2019 episode with Daniel Kahneman (1934–2024), he explained how we seize an impression and then spend the rest of the meeting defending it instead of testing it.

[sayanthan @sayanthan]

I posted sometime back about the Living Book concept, An alpha version of the Living Book platform is now live and so is the first book on it: AI Native Leap! Give it a try and share your feedback about the platform and book!

https://t.co/lkk2jZR5i5

[sayanthan @sayanthan]

@YarlITHub's journey is powered by 1000s of volunteers, partners, and sponsors who give their time, energy, and resources. A true community effort - ஊர் கூடி இழுக்கும் தேர். Big thanks to @Shenelle_r for creating this amazing video that shares this story with the world.

[Shenelle Rodrigo @Shenelle_r]

Jaffna is rising — not with skyscrapers, but with empowered youth. 🇱🇰

For 13 years, the Yarl Geek Challenge has been volunteer-run, bridging Sri Lanka’s diaspora & local youth to drive innovation and hope.

Watch how a war-torn region is being rebuilt — from the ground up.👇
🎥 https://t.co/1wX5Hbiahx

[Farnam Street @farnamstreet]

Andy Grove knew emotional attachment to past decisions could be fatal. He forced himself and his company to confront reality before it confronted them:

[sayanthan @sayanthan]

Manja AI - Sales Gym: Train like a top closer just had its product hunt launch - if you like it share and upvote it! https://t.co/eCVfb1dVRW via @producthunt

[sayanthan @sayanthan]

Just pushed “AI Sales Gym” to live: video demo/call drills w/ an AI buyer who matches your ICP & throws real objections. Try it out and punch holes and share feedback. Created a small video of how it works. You can watch it here. @manja_inc

[sayanthan @sayanthan]

If you want to practice and improve your sales pitch, sign up and do a demo on video with screen-sharing and see @manja_inc do its magic! If you haven't experienced a video call with AI, you may want to try it out as well! Its a preview, so please do share feedback!

[Manja Inc @manja_inc]

Stop Practising on Prospects. Meet Manja AI's Sales Gym! You wouldn’t run a marathon without training…Yet sales reps “wing it” 6× a day. Today we launch AI Sales Gym. Sign up and practice your pitches! https://t.co/2m2J3rW9ik

[Orange Book 🍊📖 @orangebook]

You suddenly become much more successful once you finally apply all the advice you've been giving other people.

[Adam Grant @AdamMGrant]

Procrastination is not a time management problem. It’s an emotion management problem.

You’re not avoiding work. You’re avoiding tasks that feel boring, frustrating, or stressful.

Postponing prolongs pain. The dread of starting is quickly outweighed by the joy of progressing.

[Yarl IT Hub @YarlITHub]

Unlock the Secrets of Successful Business Models! 🚀

Join us at Yarl IT Hub Mannar for an interactive workshop on Business Models – explore various strategies, gain insights into innovative business frameworks, and take your entrepreneurial journey to the next level!

[Paul Graham @paulg]

Cruft breeds cruft.

(If you're not familiar with the word "cruft," it means an accumulation of unnecessary stuff.)

[Orange Book 🍊📖 @orangebook]

The simplest way to attract interesting people into your life is to work on an interestingly hard problem that you are uniquely well-positioned to solve, that constantly forces you to upgrade and renew who you are.

[Sebastian Raschka @rasbt]

Maybe a hot take, but what about the following advice to the next gen:
Don't get an AI degree; the curriculum will be outdated before you graduate. Instead, study math, stats, or physics as your foundation, and stay current with AI through code-focused books, blogs, and papers.

[Sundar Pichai @sundarpichai]

Introducing Willow, our new state-of-the-art quantum computing chip with a breakthrough that can reduce errors exponentially as we scale up using more qubits, cracking a 30-year challenge in the field. In benchmark tests, Willow solved a standard computation in <5 mins that would take a leading supercomputer over 10^25 years, far beyond the age of the universe(!).