guo_flame_scheme

郭詐藥自燃，众郭蛾扑火，照亮了洼地生态里的愚患！一燈能照千年暗，一賤能显萬年愚…. 归纳了郭文贵武賤海外华人的事实…. 郭吞一堆藥，蛾以幻為真！ 郭的口蛊术在美帝面前就像是…… 炼丹术詐元素周期表，自取郭耻智辱！

[ali @waterloo_intern]

I spent 48 hours with the Kimi K3 modeling code.

It took:
- 650 mg of caffeine (mandatory)
- 40 cans of LaCroix (optional... world record (?))
- 8 papers
- 6 months off my lifespan

Finally grokked the entire lineage of Kimi K3 and how we got here... every single step, since 2019 GPT-2

[Calc Consulting @CalcCon]

Yurii Nesterov is a Belgian-Ukrainian mathematician and one of the founders of modern optimization theory. His work has had an enormous impact on machine learning, artificial intelligence, engineering, operations research, economics, and scientific computing.

He just recently received the Carl Friedrich Gauss Prize, one of the highest honors in applied mathematics, recognizing mathematical work that has had exceptional real-world impact.

His work laid the foundations to today's NN optimizers (Adam, AdamW, RMSProp, etc.) While the not identical to Nesterov's algorithm, many build directly on ideas that he introduced. Epecially acceleration and momentum. His work fundamentally changed how we think about designing optimization algorithms.

[Moysei @0xMoysei]

Richard Feynman was asked in the 1980s whether machines will ever think. He answered in front of a small room, on camera, decades before anyone shipped a chatbot.

Feynman won the 1965 Nobel Prize in physics. He also helped design the first parallel computers at Thinking Machines, so this isn't a physicist guessing about software.

Every AI argument running in 2026 is a rerun of the question he takes apart here. He refuses the framing most people still use and rebuilds it from the machine up.

Watch the section where he compares what machines do to what animals do. He doesn't answer yes or no. He shows why the question is shaped wrong, in plain sentences, no math.

An AI researcher I know played it for his team before a roadmap meeting and killed 2 features by lunch.

The footage sits free on YouTube, blurry, unedited.

We built the machines. We still argue with a dead man's framing.

[Epic Maps 🗺️ @theepicmap]

China has planted more trees than any other country. Since 1978, China has planted 66 billion trees to hold back two deserts.

[JTheretohelp1 @JTheretohelp1]

10- and 30-year Yields, and THE most important chart in Financial Markets…

“While much of the attention by market participants remains focused on how high equities can climb, a historic shift has occurred in the bond market. After more than 40 years of disinflation and falling yields, both the 10-year and 30-year yields have broken out from a multi-decade price channel. While a near-term risk-off event could temporarily force yields lower, the technical indicators and broader economic realities suggest this breakout will have profound implications for years to come.

Given the unprecedented level of debt accumulated by the United States, this structural shift in yields is becoming the single most critical factor shaping our fiscal future. It will directly dictate how much capital must be directed toward debt servicing rather than social programs, defense spending, and other economic growth initiatives. The ongoing lack of fiscal discipline and continued government spending come at the direct expense of hardworking taxpayers and future generations, setting the stage for a severe sovereign debt crisis that extends far beyond standard stock market volatility.

To help illustrate this technical breakout, I have attached a chart from Advisor Perspectives that I have personally annotated. This visual highlights why higher yields heading into the next decade are no longer a matter of debate, but rather a matter of time.”-JT, 7th Key Financial

[Nathie @nathie]

This robot arm lets you go base jumping in VR.

[مُضَر @almsaodii]

هل تعلم أن آخر دولة إستطاع المغول دخولها هي الصّين (سُلالة سونغ) رغم كونها أقرب دولة جغرافياً!

[DaVinci @BiancoDavinci]

This is a beautiful table.

[𝐀𝐔𝐍𝐓𝐘 𝐀𝐃𝐀 @Auntyadaa]

Son used Ai to turn his mother’s old photos Into a 60th birthday gift 💝

[Francisco Ribeiro @fraveris]

Orchard Awakening

By Jeff Stanford, 2026

[Nautilus Research @NautilusCap]

SPX Election cycle seasonal composite.

[11:11 @2025pics]

California's Highway

[Linda Raschke @LindaRaschke]

Easy read for those less familiar with Bayes theorem. Bayesian Principles are a key driver in my trading methodology.

[Here is Shanxi! @weib1591]

猜猜这是中国哪里？

[Amazing Things @amazingthings_]

A mountain range stopping clouds like a natural wall

[guo_flame_scheme @tdgscheme]

@VolumeLeaders You are the goat! Thank you!

[Goshawk Trades @GoshawkTrades]

The Livermore System

From: Trading Systems and Methods by Perry J. Kaufman

[Math Files @Math_files]

When physicist Richard Feynman visited Brazil, he once ate alone at a restaurant during off-hours. A Japanese street vendor selling abacuses came in and challenged people to do calculations faster than him. The waiters sent him to Feynman, the only customer.

Feynman lost at addition and narrowly lost at multiplication. During division, they tied. Then the vendor proposed a harder test: finding the cube root of 1729.03.

While the vendor worked furiously on his abacus, Feynman calmly thought. Knowing that , he reasoned the answer must be slightly more than 12. Using a simple approximation rule, he quickly wrote several decimal places.

The vendor could only reach 12.0 and left defeated. Feynman won by insight, not speed, using estimation and mathematical reasoning rather than raw calculation.

[Haider. @haider1]

Mathematician Terence Tao:

Training and running LLMs isn't mathematically difficult; any math undergrad could understand the basics

The mystery is that we have no theory to predict why models excel at certain tasks and fail at others

"we can only make empirical experiments"

[Gary Black @garyblack00]

@TheTeslaBull I heard the same thing in 2020-2021 when I pooh-poohed the notion that $TSLA would sell 20M EVs/year by 2030. TSLA will end 2025 with sales of around 1.65M, down for a second consecutive year.