James Hill O'Brien

I am...

[Saboteamos @saboteamos]

Compas de ex.vendetta (IG) lanzan https://t.co/PXX07hBtp3, una plataforma abierta que permite mapear y visualizar la economía del genocidio. Aún está en fase de desarrollo, pero ya está la huella de muchos gobiernos, empresas y políticos cómplices
12/10/26

[Nick Touran @whatisnuclear]

I believe this is the first-ever documentary about nuclear fusion. It was produced in 1958 just as Project Sherwood was declassified. Now I'm happy to make it available online for the first time ever. Please enjoy!

Thanks to Adam Rizika (funded scanning the 1st part, in high res) and thanks to the IAEA Lise Meitner Library (digitized the 2nd part in DVD quality)

Contents:
- 0:00 Introduction: Fuel in the Oceans and the Fourth Conquest of the Atom
- 2:15 What Is a Plasma? The Deuterium Fusion Reactions
- 4:20 Magnetic Confinement: The Magnetic Bottle, Leakage, and Instability
- 6:18 Project Sherwood's Four Labs; Los Alamos and the Perhapsatron
- 10:27 Columbus II and Columbus S4: Fast Pinch Experiments
- 12:28 Ixion and Scylla: Rotating Plasma and Fast Compression
- 15:30 UC Radiation Lab: Gamma Pinch, Homopolar, and Tubular Pinch
- 20:31 The Magnetic Mirror Machine and "Table Top"
- 24:36 "Toy Top": Multi-Stage Magnetic Compression
- 27:20 "Felix": Injecting and Trapping High-Energy Ions
- 29:30 Princeton's Project Matterhorn: The Stellarator
- 32:35 Stellarator Instability, Shear Stabilization, and Ohmic Heating
- 35:17 The Diverter and the Model C Stellarator
- 37:47 Oak Ridge National Laboratory: The DCX Machine
- 40:12 Inside DCX: Trapping the Ion Beam, Charge Exchange, and Burnout
- 44:18 DCX in Operation and the State of Fusion Research, 1958

[Michael Adams @m_adams]

Introducing the US Gov Graph - a complete map of the people and positions of power in the federal government.

To fix our institutions we must understand how they work, so we are using AI to model and monitor every government in America. This is Palantir for The People.

Check it out here: https://t.co/bRXZfEt8ud

We exist to set the conditions for an era of civic excellence unmatched since the Founding. We will need nothing less to navigate the age of AI.

I'm looking for exceptional engineers and designers to work with me in San Francisco. If that's you, inquire here: https://t.co/8hpHt2MhVu. Self-governing civilization is counting on you.

[Financelot @FinanceLancelot]

The U.S. 10 year yield is getting squeezed between 2 parabolic trendlines

The trend breakout in November will determine if we're in the 1970s or 2008.

[Jacob Coxon @hilbertspaess]

I resigned from Anthropic today. I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives. More thoughts below.

[Avery Tomasco @averytomascowx]

Incredible hi-res satellite loop of the deadly supercell that crashed into Central Texas yesterday evening.

Hail up to 2" diameter locally... Microburst winds 70-90 MPH in Austin... Rainfall rate over 10" per hour with one flood-related fatality. Terrible storm #atxwx #txwx

[TravisCountySevereWx @TravisCOSW]

Heads up, friends. Storms moving thru Austin (esp E Austin) could be strong with with wind gusts up to 50mph (58 is severe threshold) & small hail #atxwx #txwx

KEWX - Super-Res Reflectivity 1, 6:02 PM

[US Homeland Security News @defense_civil25]

Storms just came through Austin and Central Texas. It sounded like a small hurricane for about 20 minutes. It scared the crap out of my dogs!! Stay safe out there.

[Allegra Jacchia @allegrajacchia]

Today, we’re launching Original Content Rewards.

The reality is that Revenue Sharing had reached a point where its incentives were misaligned. Creators should be focused on bringing net new content to the platform instead of maximizing payouts. We could have kept adding more rules and exceptions, but ultimately the better decision was to start fresh and build a program designed from day one to reward originality.

Original Content Rewards is the result.

This program is built around a simple idea: reward the creators who bring original ideas, expertise, creativity, and unique perspectives to 𝕏 -- not those who have become best at gaming the system.

This is just the beginning. We’ll continue refining the program, improving our models, and raising the bar over time. But we’re so excited to keep investing in the creators who inform us, entertain us, make us laugh, teach us something new, and drive meaningful conversations on 𝕏 every day.

To everyone who creates original content on 𝕏 and makes this a better platform every day: thank you. We’re excited to keep building for you.

[Noam Brown @polynoamial]

An internal version of Astra, @OpenAI’s next major model family, solved 10 major open problems in mathematics, quantum complexity, and theoretical computer science.

We believe it will be a major step for scientific reasoning. https://t.co/iP6cyheZ7i

[Doug Lewin @douglewinenergy]

Air conditioning demand in extreme Texas heat pushed the ERCOT grid beyond 91 gigawatts today. Before yesterday, we had never exceeded 86 GW so this record shatters the old one. Note there was >20 GW of buffer, no energy emergency, no conservation call. #txlege #txenergy 1/

[Superman @thesupermannx]

China open-sourced a peanut-sized OCR that parses entire 100-page PDFs in one shot..

It's called Unlimited-OCR. Only 3B params. Runs locally.

Every other OCR tool chops your doc into pages and loses the thread. this one reads the whole thing in a single pass.

→ One-shot "long-horizon" parsing (32K context window)
→ Multilingual, out of the box
→ 93% on the standard parsing benchmark (+6 over baseline)
→ <0.11 error rate past 40 pages
→ Runs 100% locally on your own hardware
→ Works with Transformers, vLLM, SGLang, Docker, Ollama, llama.cpp

Traditional cloud OCR (Textract, Google Vision, Azure Doc Intelligence) costs $1.50–$15 per 1,000 pages.

This runs on your machine. For free. Forever.

Baidu built it explicitly to push DeepSeek-OCR one step further. Already at 1.9M downloads on Hugging Face and most people have no idea it exists yet.

100% open source.

[Anatoli Kopadze @AnatoliKopadze]

The engineer who built Claude Code just dropped a 28-minute video on how to write prompts that actually work.

I've seen $300 courses that don't cover what he shows in the first 10 minutes.

CLAUDE.md files, memory shortcuts, parallel sessions, prompting patterns.

Watch it, then read the guide below on the Claude features 99% of users never find.

[CyrilXBT @cyrilXBT]

https://t.co/lkgDX3DkdY

[Anatoli Kopadze @AnatoliKopadze]

https://t.co/eq0oL4fIaR

[Massimo @Rainmaker1973]

Scientists have created one of the most detailed 3D reconstructions of a human cell (eukaryotic cell) ever produced.

This groundbreaking model, often termed a "Cellular Landscape Cross-Section Through a Eukaryotic Cell," combines data from X-ray tomography, nuclear magnetic resonance (NMR), and cryo-electron microscopy to map molecular structures in extreme detail.

[neurofiber @neurofiber]

@Rainmaker1973 This zoomable chart of the Roche Biochemical Pathways shows the metabolic processes that make Eukaryotic Cells function.
https://t.co/kzGg76nkoI

[Alpha Insights @AlphaInsightsAI]

Terafab: Elon Musk’s Ambitious Plan to Build a Massive AI Chip Factory

Terafab is one of the most ambitious private semiconductor projects in recent history. Announced in March 2026, it is a joint initiative between Tesla, SpaceX, and xAI, with Intel joining as a key technology partner. The project aims to create a highly vertically integrated chip manufacturing facility capable of producing advanced AI processors at an unprecedented scale.

Why Terafab Is Being Built

Tesla, Optimus humanoid robots, xAI’s Grok models, and SpaceX’s future orbital infrastructure all require enormous quantities of custom AI chips. Elon Musk has argued that existing chipmakers cannot keep up with the projected demand. Terafab is designed to give these companies greater control over their silicon supply, reduce dependency on external foundries, and dramatically speed up the chip development cycle through vertical integration.

What Makes Terafab Unique

Unlike traditional chip fabs, Terafab is planned as a highly integrated operation. The goal is to bring together logic chip fabrication, memory production, advanced packaging, testing, and even lithography mask production under one roof or campus. This approach would allow engineers to design, build, test, and refine chips in a continuous loop without the delays of shipping wafers between different specialized facilities around the world.

Location

The project has two main components:
1. A prototype and research fab is already under construction at the North Campus of Giga Texas in Austin. This smaller facility will focus on rapid iteration and testing new processes.
2. The much larger full-scale Terafab is expected to be built at a separate site, with Grimes County, Texas, frequently mentioned as a likely location.

Timeline

- March 2026: Official announcement of the project.
- 2026: Construction of the prototype fab at Giga Texas is already underway. Small-batch production of the AI5 chip is targeted for late 2026.
- 2027: Volume production is expected to begin at the research fab.
- 2027–2028 and beyond: Phased ramp-up of the full-scale facility.
- 2030–2032: Full high-volume production capacity is anticipated in the most optimistic scenarios.

Investment and Costs

The financial scale is significant:
- The research/prototype fab at Giga Texas is expected to cost around $3 billion.
- The overall Terafab project is estimated at $20–25 billion for the core initiative.
- Longer-term and expanded phases could push total investment much higher, with some internal estimates reaching $55 billion initially and up to $119 billion across all phases.

Production Targets

At full scale, Terafab’s ambitions are enormous:
- Wafer production capacity targets start at 100,000 wafer starts per month, with a long-term goal of up to 1 million wafer starts per month.
- Annual output is projected to reach 100–200 billion custom AI and memory chips.
- The ultimate goal is to deliver more than 1 terawatt of AI computing power per year.

The facility will focus on advanced process nodes (targeting 2nm-class technology), with Intel’s 14A process expected to play a role in scaled production.

[CyrilXBT @cyrilXBT]

MIT just quietly dropped a free AI curriculum that puts $50,000 university courses to shame.

12 books.

Zero tuition.

From the same institution that produced the people building the models everyone is talking about.

FOUNDATIONS

1. Foundations of Machine Learning — https://t.co/Un6UbjJBMW
2. Understanding Deep Learning — https://t.co/UQxZmyFq2V
3. Machine Learning Systems — https://t.co/YAgrLVH8N1

ADVANCED TECHNIQUES

4. Algorithms for ML — https://t.co/YlBk59oGwX
5. Deep Learning — https://t.co/KMO1uWQ69z

REINFORCEMENT LEARNING

6. RL Basics (Sutton & Barto) — https://t.co/sOZlDXA1Tz
7. Distributional RL — https://t.co/uOkviYj8fF
8. Multi-Agent Systems — https://t.co/Dx9caJW4QL
9. Long Game AI — https://t.co/K9Qm2Tk8FE

ETHICS & PROBABILITY

10. Fairness in ML — https://t.co/MgkLdRvQ2m
11. Probabilistic ML Part 1 — https://t.co/Zz33gQizle
12. Probabilistic ML Part 2 — https://t.co/qBe776ERrO

This is a complete MIT-level AI education.

Not a YouTube playlist.

Not a Twitter thread full of fluff.

Textbooks written by the researchers who built the field.

The people who actually study this will not just understand AI better than their peers.

They will understand it better than most people currently getting paid to work in it.

Most people will bookmark this and never open it.

The ones who open it tonight are the ones who show up in 12 months having built something nobody around them understands yet.

Bookmark this.

Open the first one tonight.

Follow @cyrilXBT for more resources that actually compound.

[zerohedge @zerohedge]

Next: attention turns to the $1.8TN in shady accounting hidden liabilities (thanks SPVs):
$1TN in purchase commitments
$0.8TN in leases not yet started
... oh and $2TN in RPOs and throw in a little chip depreciation