TrueNorth

[KUTUB @KUTUBProduction]

Israel flattened Beirut in 1982.

No Hamas. No Hezbollah. No October 7 to point to then.

Just 17,000 dead Lebanese and Palestinian civilians.

Even US president then Ronald Reagan, who armed Israel, called Begin furious after seeing a photo of a 7-month-old baby with its arms blown off and said “It is a holocaust”.

They killed so many innocent people that the survivors had no choice but to pick up weapons.

Then Israel had the audacity to keep using “Self-Defense” excuse every decade.

Israel didn’t stumble into endless war. Israel built it. Brick by brick.

Own it.

[rewind @rewind02]

A Stanford professor just gave a public lecture on exactly how GPT, Claude, and LLaMA are built under the hood

no insider access required

just the clearest breakdown of modern LLM architecture I've seen

this lecture reveals the framework professors are paid up to $750K a year to teach

the gap between "I use ChatGPT" and "I understand how it works" is smaller than most people think

the most complete public breakdown of modern LLM architecture I've seen this year

[Nadira Ali🇵🇸 @Nadira_ali12]

Israel destroyed the mausoleum of Simon Peter, apostle of Christ, in the village of Shama in southern Lebanon

1,925–1,995 years years old .

Christians of the world, Wake up!!

[RT @RT_com]

Blood pours from Trump's birthday cake as Iranian state media releases AI video showing the 168 children killed in Minab school strike returning to haunt him on his birthday

'You gave the order to strike our school, exactly when all my friends were sitting in their classrooms'

[zaimiri @zaimiri]

https://t.co/a8ejOfyUQr

[Quant Science @quantscience_]

How to make a simple algorithmic trading strategy with a 472% return using Python.

A thread. 🧵

[CyrilXBT @cyrilXBT]

this is f*cking gold

How to build your own AI agents that actually work in the real world - full course

if I had this a year ago, I would've shipped my first agent in a day instead of 2 weeks

in the right hands, this changes everything:

[Ren @ren_stocks]

Satya Nadella just posted something that validates the entire AI buildout thesis from the very top of the stack.

The model is commoditizing. The durable value is the learning loop a company builds on top of the model.

He splits it into two assets:

Human capital -- the knowledge, judgment, relationships, and pattern recognition of your people.

Token capital -- the AI capability the firm builds and owns.

He says the real opportunity is building a learning loop where human capital and token capital compound together.

If the model layer is commoditizing then the durable returns are not in the model makers. They are in the infrastructure that powers every company building its own loop. Compute. Memory. Interconnect. Power.

The full stack underneath the application layer.

The model wars will have winners and losers. The infrastructure underneath gets bought either way.

Bullish the AI buildout.

Every layer. If you want to understand them in detail, check out my Substack.

https://t.co/Wna5UzCOVT

[Mr. Buzzoni @polydao]

GOOGLE DEEPMIND'S DIRECTOR OF RESEARCH JUST SHOWED WHY CLAUDE AND EVERY OTHER LLM IS STILL THINKING SLOW

Brendan O'Donoghue on text diffusion - the architecture that generates entire blocks of text at once instead of one token at a time

this video answers:

> why current LLMs hit a speed ceiling no matter how good the model gets
> how a model can write the wrong answer first, finish its reasoning, then go back and fix it
> what 2000 tokens per second actually looks like in a live demo
> why this runs on your phone before it runs on a server

while everyone was debating Claude Opus 4.8 vs Fable 5 - which the US blocked last week - O'Donoghue presented the architecture that makes the comparison beside the point

save this 👇

[Avi Chawla @_avichawla]

The ultimate Full-stack AI Engineering roadmap to go from 0 to 100.

Bookmark this.

This is the exact mapped-out path on what it actually takes to go from Beginner → full-stack AI engineer.

> Start with coding fundamentals.
> Learn Python, Bash, Git, and testing.
> Every strong AI engineer starts with fundamentals.

> Learn how to interact with models by understanding LLM APIs.
> This will teach you structured outputs, caching, system prompts, etc.

> APIs are great, but raw LLMs still need the latest info to be effective.
> Learn how LLMs are usually augmented with more info/patterns.
> This will teach you the basics of fine-tuning, RAG, prompt/context engineering, etc.

> Strong LLMs are useless without context. That’s where Retrieval techniques help.
> Learn about vector DBs, hybrid retrieval, indexing strategies, etc.

> Once retrieval is solid, move into RAG.
> Learn to build retrieval + generation pipelines, reranking, and multi-step retrieval using popular orchestration frameworks.

> Now, step into AI Agents, where AI moves from answering to acting.
> Learn memory, multi-agent systems, human-in-the-loop design, Agentic patterns, etc.

> Learn how to ship in production with Infrastructure.
> This will teach you CI/CD, containers, model routing, Kubernetes, and deployment at scale.

> Focus on observability & evaluation.
> Learn how to create eval datasets, LLM-as-a-judge, tracing, instrumentation, and continuous evaluation pipelines.

> Security is crucial.
> Learn how to implement guardrails, sandboxing, prompt injection defenses, and ethical guidelines.

> Finally, explore advanced workflows.
> This covers voice & vision agents, CLI agents, robotics, agent swarms, and self-refining AI systems.

This is the actual journey to becoming a full-stack AI Engineer and not just "use” AI, but designing full-stack AI systems that can survive in production.

If you need specific resources, I wrote a detailed article that provides a structured learning roadmap for AI engineers in 2026.

It covers prompting, RAG, fine-tuning, agents, MCP, evals, and inference, with guidance on what to prioritize and in what order.

Read it below.

[Rahul @sairahul1]

This is the best site on the internet to learn how LLMs actually work.

Free. Completely.

https://t.co/YOGF6PsmBN

Bookmark this site.

Then read this setup ↓

[Roan @RohOnChain]

Probably the only lecture you need to master "game theory".

Bookmark now, learning game theory can change your life btw.

[Michael McDonough @M_McDonough]

🚢Claims about Hormuz reopening are exactly why we built this:

If ships start moving through the Strait with AIS transponders on, you’ll see it in the index — updated every 30 minutes alongside Suez, Bab al-Mandeb and Panama.

[Ruuj @RuujSs]

https://t.co/JsXybTHUpJ

[Unsloth AI @UnslothAI]

You can now run Kimi K2.7 Code locally! 🌘

We shrank the 1T model to 325GB (-48%) via Dynamic 2-bit where important layers are upcasted.

Run at >40 tok/s on 330GB RAM/VRAM setups.

Run full precision on 610 GB.

Guide: https://t.co/SXZJ3IHMpY
GGUF: https://t.co/2lpUx7u0r8

[Curiosity @CuriosityonX]

🚨: MIT has made almost all of its courses available for free online — allowing anyone to learn for absolutely no cost

[Quant Science @quantscience_]

Using Mathematics to Make Money by legend Jim Simons.

8 page PDF free for download:

[How To Prompt @HowToPrompt__]

Researchers open-sourced an AI that taught itself 300 years of physics with zero physics knowledge

In 1907, Einstein had what he called his "happiest thought": gravitational mass = inertial mass. It took him 8 more years to turn that insight into General Relativity.

Now, researchers at Peking University built an AI that figure out the same thing on its own with zero physics knowledge.

They call it AI-Newton.

They didn't train it on physics textbooks. They didn't pre-program any formulas.

They just fed it raw, noisy experimental data and let it explore.

The AI started defining its own concepts.

First, it measured the stretch of a spring hanging from a weight. It invented the concept of "gravitational mass."

Then, it measured the oscillation frequency of a bouncing spring. It invented the concept of "inertial mass."

And then, entirely on its own, the AI noticed the numerical equivalence.

It realized these two completely different physical measurements were fundamentally the exact same thing.

It merged the concepts. It had Einstein's realization.

But it didn't stop there.

It systematically went on to autonomously rediscover Newton's second law, the conservation of energy, and the law of universal gravitation.

We have spent decades using AI to crunch numbers for human scientists.

But this is a paradigm shift.

AI is formulating concepts. It is making the intuitive leaps that we thought only belonged to human genius.

If an AI can rediscover the foundations of modern physics by just looking at raw data...

What is it going to discover that we haven't thought of yet?

[S p r i n t e r @SprinterPress]

🇮🇷 IRAN HAS WON. THE USA HAS BEEN HUMILIATED.

Three months ago, Iran was considered doomed. Its top leadership had been annihilated, its air defense system was practically non-existent, and its economy was collapsing. The enemy was using the latest AI technology, the country was swarming with spies, public support was wavering, and the opposition was rebelling. It seemed that defeat was inevitable.

Today, Iran is dictating terms. The USA is forced to make a humiliating compromise.

How did this happen? Not through numbers or technology. But through firm principles.

Iran doesn't play by the rules. It strikes where it hurts most, without regard for conventions. It has genuine convictions – ideas that the authorities and elites are willing to defend at the cost of their own lives. Iran doesn't beg for negotiations – it's the USA that came to it.

US technology, economic superiority, and intelligence don't guarantee victory if there's a willingness to launch asymmetric strikes. Iran has proven that even in an almost hopeless situation, you can turn the tide of the game.

💯 War with the USA and the West is not a verdict, if there's willpower and systematic work on mistakes.

[Lyn Alden @LynAldenContact]

If you denominate US GDP in gold instead of dollars, the chart is wild.