DMX

Conservative liberal in the sense of Milton Friedman

[DMX @2010dmx]

@CalltoActivism he is such a weasel

[DMX @2010dmx]

@jianmang054A 是的，只能说你也变成了何不食肉糜的那种站着说话不腰疼的人罢了

[灰狐 @huihoo]

鸢尾花书家族又添新书
”数学不难“ 之 《线性代数不难》上下册
https://t.co/T05YAXjWM5
油管也有32节课视频，可学习参考
https://t.co/a0nVX1DPqx

[Kirk Borne @KirkDBorne]

Introduction to Modern Statistics: https://t.co/COq0CLz1Ds
📊📈
Choose:
1️⃣Read it FREE online.
2️⃣Or name-your-own-price for the PDF.
3️⃣Or buy the book here: https://t.co/wiF6qwHL09
————
#Mathematics #DataScience #DataScientist

[Tech with Mak @techNmak]

A University of Colorado Boulder professor built one of the best AI-learning resources I’ve seen for people who are tired of learning about models through diagrams and code alone.

Prof. Tom Yeh makes students calculate modern AI architectures by hand.

Transformer. Self-attention. Multi-head attention. MoE. Switch Transformer. Mamba S6. CLIP. U-Net. Sparse Autoencoders. BitNet. RLHF. Even Diffusion Transformers.

The idea is simple => shrink the architecture until the full forward calculation fits on a page, then work through every number with a pen. Yeh describes the project as reducing each model to an example small enough to calculate end-to-end by hand.

Learn here: https://t.co/DXfVjhdGkl

[空空道人 @Kongkongda5882]

说个魔幻的事。
​一台在国内被强制报废、连年检都过不了的重卡，拉到非洲，能被当成"机皇"抢着要。
​而把这事儿干成产业链的，不是北上广深，是山东一个连高铁都没通的小县城——梁山。
​2025年，这里卖出去了1.3万辆二手商用车，全国县域第一。今年上半年光往非洲就塞了7000多台，济宁一个市的出口量同比涨了七成。
​非洲兄弟不傻，他们算过账。
​欧洲二手重卡贵得离谱，配件还得等海运，动不动趴窝半个月。中国淘汰的车呢？价格只有三分之一，配件满大街都是，随便找个路边摊拿扳手就能敲。关键是国内觉得"废了"的国三国四车，在非洲刚过"磨合期"——那边路况烂、超载狠，反而就吃这种粗粮胃口好的老机器。
​但最让我服气的，是梁山人的玩法。
​二手车出口最大的雷是车况。你收一堆跑了几十万公里的破烂，洗洗就发出去，车在尼日利亚半路趴窝，人家下次还信你？
​梁山这帮人直接把"淘汰"和"报废"划清界限。底盘开裂的不要，缸体有裂纹的不要，重大事故车碰都不碰。收回来进再制造中心，发动机整个拆散，清洗、探伤、换磨损件，按新机器的标准重新装配，跑完热磨合才算完。
​底盘要调、变速箱要验、实车要路试、整车要淋水——四关全过才给装柜。
​硬生生把"废铁价"的车，整备出了"新车底气"。
​更狠的是效率。以前"一车一证"，几十台车做几十套单子，手续跑两周。现在改成"一批一证"，三天拿证。海关设了二手车专岗，7×24小时预约通关，单车过关十分钟。
​海外售后也没糊弄。尼日利亚、肯尼亚直接建自有仓库和维修点，出货量小的国家就靠当地经销商兜底。目前梁山企业在非洲已经建了7家海外工厂。
​一个鲁西南小县城，没港口没边境，硬是靠着一堆别人不要的铁疙瘩，把自己干成了全球基建的"备胎仓库"。
​不吹不黑，这买卖干得漂亮

[李老师不是你老师 @whyyoutouzhele]

后续：

8月17日，联合早报报道，浙江杭州滨江区副区长郁廷栋简历被政府官网撤下后，一篇和他有关的微信公众号文章《一场酒局献祭，掀开了杭州地产潜规则》在中国网络热传。该文作者据称是前调查记者。

[鸟哥 | 蓝鸟会🕊️ @NFTCPS]

李笑来推荐的那本《研究是一门艺术》，说是AI时代必读，我看了下，还真不是随便吹的。

给你捋清楚版本这事：
1️⃣ 英文版都出到第五版了
2️⃣ 中文电子版是照着第四版翻的，2021年出的，中文这边就这一个最新的
3️⃣ 纸质版也就30块，一顿饭钱都不到，真想读别抠这点

我自己的判断是，AI越铺开，以后不光有一人公司，还会冒出一堆一人研究者。工具到位了，剩下就看你会不会自己找问题、自己钻。

阅读链接：🔗https://t.co/Rali9E6Rf5

[Kirk Borne @KirkDBorne]

Free Mathematics PDF —  MIT guide to the mathematical foundations of computer science: https://t.co/XCLY3DAs9w [1048 pages]

[tetsuo @tetsuoai]

In C, `array[i]` does not add i to the address. It adds i times the size of one element. Assembly has to do that multiply. C hides it.

Stanford CS107, lecture 9.

[Kirk Borne @KirkDBorne]

Introduction to Graph Theory: https://t.co/wP0nfijgIf (454-page PDF) — edges, vertices, trees, networks, paths, and graphs (and theorems!)
—————
#Mathematics #NetworkScience

[Mustafa @oprydai]

most programmers learn math backwards.
theory first. application somewhere later.

math for programmers by Paul orland does the opposite.

→ vectors become movement in 3d space
→ matrices become transformations
→ calculus becomes motion and optimization
→ probability becomes reasoning under uncertainty
→ linear algebra becomes the machinery behind machine learning

then you implement the ideas in python.

this is how math should be learned for engineering:
concept → intuition → code → experiment

you stop seeing equations as symbols to manipulate.
you start seeing them as executable descriptions of reality.

[Antonio Lupetti @antoniolupetti]

"Fundamentals of Linear Algebra" is an excellent 400-page textbook on linear algebra, freely available from the University of British Columbia.

It covers linear systems, matrices, vector spaces, linear transformations, determinants, eigenvalues and eigenvectors, inner product spaces, diagonalization, and much more. It also includes some interesting applications, including linear coding theory and many exercises.

I think it is a very useful resource if you are studying linear algebra and want a substantial reference to keep at hand.

https://t.co/oWytdLxTaN

[Antonio Lupetti @antoniolupetti]

My little essay on "Learning Mathematics" has received a lot of attention, and I'm really happy with all the feedback I've received. If you missed it, I'm sharing it again here. I'm already working on the second instalment.

[Andrej Karpathy @karpathy]

Agency > Intelligence

I had this intuitively wrong for decades, I think due to a pervasive cultural veneration of intelligence, various entertainment/media, obsession with IQ etc. Agency is significantly more powerful and significantly more scarce. Are you hiring for agency? Are we educating for agency? Are you acting as if you had 10X agency?

Grok explanation is ~close:

“Agency, as a personality trait, refers to an individual's capacity to take initiative, make decisions, and exert control over their actions and environment. It’s about being proactive rather than reactive—someone with high agency doesn’t just let life happen to them; they shape it. Think of it as a blend of self-efficacy, determination, and a sense of ownership over one’s path.

People with strong agency tend to set goals and pursue them with confidence, even in the face of obstacles. They’re the type to say, “I’ll figure it out,” and then actually do it. On the flip side, someone low in agency might feel more like a passenger in their own life, waiting for external forces—like luck, other people, or circumstances—to dictate what happens next.

It’s not quite the same as assertiveness or ambition, though it can overlap. Agency is quieter, more internal—it’s the belief that you *can* act, paired with the will to follow through. Psychologists often tie it to concepts like locus of control: high-agency folks lean toward an internal locus, feeling they steer their fate, while low-agency folks might lean external, seeing life as something that happens *to* them.”

[Furiokao @furiokao]

A dead MIT professor watched generations of engineers fall in love with the wrong thing. Neural nets. Genetic algorithms. Bayesian probability. Every year, a new method. Every year, the same blind spot.

Patrick Winston ran the MIT AI Lab for twenty-five years. He died in 2019. He gave it a name: mechanism envy.

Falling in love with a tool before asking what the problem is.

He taught the fix for fifty years. Start with the competence you want to understand. Find a representation that exposes the constraints. Only then pick a method. Without the right representation, you are guessing with expensive machinery.

Wall Street has the same disease. A trader buys a Bloomberg terminal before learning expected value. A founder hires a data team before understanding base rates. A retail investor opens a brokerage app before knowing that the number on the fund brochure is not the number that hits their account.

Quant firms do not pay half a million for methods. They pay for five representations: expected value, base rates, ergodicity, conditional probability, signal versus noise. Five ways of seeing that change every bet before you touch the machinery.

The lecture is free. The five models fit in one article. Almost nobody installs the representations before reaching for the tools.

The representations are free. Installing them is the entire edge.

[Zyphor @Zyphorrl]

HARVARD FILMED A PROFESSOR EXPLAIN IN 48 MINUTES WHAT TOOK MATHEMATICIANS 200 YEARS TO FIGURE OUT - AND SHOWS WHY THE SMARTEST MAN WHO EVER LIVED GOT THE INTUITION COMPLETELY WRONG

This is Joe Blitzstein, Harvard Statistics 110, lecture 4. He has won Harvard's Excellence in Teaching award multiple times, his textbook is used in over 200 universities worldwide, and his course has been taken by over 2 million people across 190 countries. He opens with one claim - conditioning is the soul of statistics. Everything else in the course follows from that.

He starts with De Montmort's matching problem. A deck of n cards labeled 1 through n, flipped one by one. What is the probability at least one card lands in its own position? The exact answer collapses into 1 minus 1 plus 1 over 2 factorial minus 1 over 3 factorial, continuing to n terms - which is exactly the Taylor series for e to the x at minus 1. The probability of no match converges to 1 over e, which is 0.37, no matter how large the deck gets.

Then the Newton-Pepys problem, 1693. At least one six from six dice, at least two sixes from twelve dice, or at least three sixes from eighteen dice - which is most likely? Pepys bet on the third. Wrote to Newton. Newton calculated correctly and showed it was the first, probability 0.665. Then the real punch - Newton's intuitive argument was wrong, and a statistician named Stigler proved it without even understanding what Newton wrote. Newton's argument never used the fact that the dice were fair. So it could not possibly be correct.

Then conditional probability. The definition is a single fraction - probability of A given B equals probability of A and B divided by probability of B. Blitzstein explains it two ways. First, pebble world - learning B occurred simply erases every outcome outside B and renormalizes what remains to sum to 1. Second, frequentist world - repeat the experiment many times, circle every run where B occurred, ask what fraction of those also had A.

Watch the moment he derives Bayes' rule in 10 seconds - divide both sides by P of B, end of proof. Then says controversies about this one line have raged for centuries and an entire field of statistics was built on top of it.

A data scientist I know rewatched this before switching careers into statistics. Said it was the first time probability felt like a system with rules rather than a collection of tricks.

Free on YouTube, Harvard, over 2 million views.

bookmark this and watch later - after this lecture every time you learn something new you will feel yourself updating a probability rather than changing your mind

[Lumen @lumenxbt]

https://t.co/3hWKMIBJcw

[DMX @2010dmx]

@Oleksan5705847 Great