Objectively Random

Algo Trading, Strat, Macro. Math PhD (Erdős 3). Hon Reader@UCL. Created MSc Algo Trading (‘17-now), & teaching online. Too much street-side experience to list.

[Objectively Random @ObjRandom]

(High Speed) Formally Verified Order Matching Engines

Provably correct code is no longer limited to just mission-critical systems. Now, formal proofs are at the heart of the revolution in mathematics and are making inroads in finance.

Sign up now:
https://t.co/Ya5UNP08vP

[Ash Jogalekar @curiouswavefn]

This review of Federer's "Geometric Measure Theory" by someone who was struggling to understand it for years is pure gold.

"It was only a couple of years ago, after writing a 1500 page book (now 2015 pages) on the foundations of differential geometry that I finally managed to get the gist of the first 8 pages."

https://t.co/yCPsxNZtY6

[Rishi Mishra @aRishisays]

some light weekend reading. h/t our friends @ObjRandom and @jnordvig

[Objectively Random @ObjRandom]

@Kathleen_Tyson_ @RobRastapop @TheEconomist Prabowo was the one who agreed to overflights. It was shelved when the backlash was too high. He has joined the “Board of peace”. Hard to say he is an indecent thinker being ex-military implicated in personally killing opposition under Suharto. He is not like Joko.

[Azfar Rizvi @AzfarRizvi]

In 2018, The Washington Post ran a piece on where AIPAC came from.

It's worth reading slowly.

Oct. 1953: Israeli forces killed more than 60 civilians in the village of Qibya.

Time reported it.

The New York Times published U.N. findings refuting Israel's lies about it.

Eisenhower froze U.S. aid.

Dulles said so openly.

An ordinary country faces consequences.

This one (Israel) built a machine.

5 months later, the American Zionist Committee for Public Affairs (AZCPA 0was launched. It was later renamed AIPAC.

The Post's own subhead says it "was formed to spin positive PR after Israeli atrocities."

Not to question the atrocity. To manage the fallout.

Then came the model:

... Build a "united front" and treat any dissent as betrayal.

... Operate on the premise that "there was nothing Israel might do that would jeopardize American Jewish support."

... Worse, the article says the more disturbing Israeli behavior got, "the more Israel needed their ardent advocacy."

... Since the late 1970s, steer campaign money toward chosen candidates, creating "a unity that politicians have had to respect."

That last line is the whole story. Politicians learned that crossing this lobby has a price. So they stopped asking hard questions about where American tax dollars and American diplomatic cover go.

That is leverage over American policy, built after a massacre, and sustained for 70 years.

Notice the Post didn't say American Jews are a monolith. It said AIPAC works hard to create the perception of one, and punishes Jews who break from it.

An American public that funds this deserves to know the origin story.

The question isn't "why is this being said now?" It's "why did we stop saying it?"

[Bernie Sanders @BernieSanders]

AIPAC is now spending as much as $9 million against Abdul El-Sayed.

AIPAC is why we’re at war in Iran.

AIPAC is why the U.S. has supported genocide in Gaza.

We must say NO to AIPAC.

YES to electing Abdul as Michigan’s next U.S. Senator.

[Objectively Random @ObjRandom]

@RabbiEliyainwm @RabbiJeffx @grok

[Hillel @hillelogram]

It is a weird weird year to be a formal methods expert.

Some unorganized thoughts on the new @AnthropicAI and @theoremlabs post on formal verification (https://t.co/YoymKXYFJa):

- This is very cool.

- The main innovation is verifying a later stage of the program. Normally provers take the source code, LLM it into a theorem language, and then prove that, which means you have to make assumptions about source semantics and compilation properties. It's very easy to make a mistake in translating and prove the wrong thing!

What they instead do is compile the code into assembly, and then (if I understand this correctly), compile that assembly into a theorem language (which they call LPS). So as long as the compiler is faithful you can skip the whole divergence problem.

- I'm glad they clarify that you still have to trust the translation process, the "hardware, OS, OCaml" compiler, the LPS model, and also assume that the program is single threaded with fixed memory. The single biggest source of errors in formally verified code is when the proof assumes something that isn't true in practice. It's really, really important to state your assumptions!

- I'm not clear why they need to compile the fuzzer into the theorem. Is it to act as a description of valid inputs and workload? Maybe to embed assertions in as theorems?

- Holy crap, 180 million lines of Rocq. Is that more Rocq than exists in the world already? I notice they say a 600,000 line proof corresponds to roughly 20k CPU instructions. That would be like, 1-2k lines of source code at most? Really goes to show how inefficient this approach is and how it only works because you can throw LLMs at the problem.

- This, probably deliberately, doesn't address the single biggest gap in AI-based verification: AIs cannot come up with good properties. Doesn't matter if you're doing unit testing, property testing, formal verification. I always assumed this was a "frontier labs haven't trained for it yet" thing, but if Anthropic hasn't talked about that here, maybe they've tried and haven't been successful yet. That's interesting to me.

Instead, they take a known universal property (this code doesn't crash or get flagged by UBsan) and prove that instead. Still very useful, but makes it harder to generalize to proving arbitrary things about arbitrary programs.

- Anybody reading the official post should also read their expert review appendix (https://t.co/qQU0E0nJwW). It points out some interesting things, like the fact that most of the code the LLM wrote wasn't used in the proof (over one billion lines of Rocq), that they had to prove the equivalence between fuzzed and unfuzzed programs, and (this is most interesting), some experiments find that the LLM-method of proofwriting may not scale to more complex things.

- Also read the description of LPS (https://t.co/hRTCL05X0H) to understand the limitations of the assembly -> Rocq compilation. In particular, it does not handle concurrency at all (which is one of the main use cases for formal methods!) or OS signals. Also it can't prove things about program arguments with slash characters, meaning they can prove stuff about `rm foo` but not `rm foo/bar`.

- The official skill (https://t.co/weDOdmJUfO) doesn't tell us much about the nature of FM and AI. Probably too much for me to hope of any skill.

- Some of the article's specific caveats and throat clearings make me think someone at Anthropic reads my stuff ( ͡° ͜ʖ ͡°)

[prof-g @prof_g]

increasingly concerned that ai is revealing the mathematical world as negatively curved. instead of everyone wandering about in idea-space and clustering around the same famous conjectures (local positive curvature), ai is pushing out the boundaries rapidly and revealing that the space of mathematical ideas is ~hyperbolic (negative curvature). though this means there's lots left to discover, it also means that small separations between people/agents will lead to wandering alone.

we may be emerging into an idea-space where there's so much to see and do but everyone is alone in a private heaven, unable to share their subtle joys with others...

[Objectively Random @ObjRandom]

@nypost Love it. 🥰

[Objectively Random @ObjRandom]

@mehdirhasan Why give airtime to that low IQ idiot. What he knows is a small subset of other commentators. Yglesias is a severely retarded man.

[Objectively Random @ObjRandom]

@TeeVeeInspector His moral clarity is exceptional. It is true. It does not detract from the misery of those who escaped the Holocaust, but it is not the same as being a Holocaust survivor.

[Objectively Random @ObjRandom]

@NewStatesman Did you mention he strangely likes Israeli gays? Consistent? Definitely not!

[Eyal Yakoby @EYakoby]

Owen Jones went to a Zionist party for gay Israelis and Jews in the UK this past weekend.

The man spends 24/7 bashing Zionism just to go and party with Israelis and listen to Israeli music.

[WquGuru @wquguru]

全世界包括清北的编程课程都应该向斯坦福的CS146S Fall 2026看齐！这门课相比一年前的Fall 2025已经砍掉了85%的内容，可以知道编程领域进展有多恐怖了，如果还停留在教科书十年不需更新的幻觉中，可想而知会多落后了！

两版课纲我逐周对了一遍：

第 1 周：教你怎么写 prompt → 200 行代码手搓一个 Claude Code
期末项目占分：80% → 50%
给真实开源项目提 PR：0 → 30%

去年还单独讲一周的 AI IDE、Warp 终端、一句 prompt 生成整个应用，今年全部都拿掉。

取而代之的是 Agent Skills、CLAUDE.md / AGENTS.md / hooks、agent-ready 代码库、后台并行 agent、MCP portal 和模型路由，最后一周单独讲软件工厂。

刚入门的话，我建议从第 1 周跟起：
9/22 那节 200 行手写 coding agent，slides 和完整代码都公开了，最核心的是 agent 原型—— read / write / edit / bash 四个工具加一个循环
9/24 拆主流 coding agent 的 system prompt，原始 prompt 文件也已放出来

如果已经天天跟 coding agent 打交道，前几周大概率扫一眼就行，我会重点盯这几块：
- 嘉宾：SpaceX SI负责人 @leerob、Claude Code之父 @bcherny 已经分享完，10/22 轮到 Factory 的 CTO @EnoReyes，后面还有 Cognition、Semgrep、Cloudflare、Replit 等的 @amasad 和 @eladgil
- 第 9–10 周：大团队怎么落地，MCP portal、LLM gateway、模型路由、成本优化
- 前五周的阅读清单，凑起来就是一份 2026 版 agent 工程书单

30% 的开源分也很有意思：Vercel、Semgrep、OpenHands、Unsloth 等 15 个开源项目给学生做指导，PR 直接往生产级代码库里提，这在大学课程里还是头一回。

课程材料陆续公开，不是斯坦福学生也能跟：https://t.co/UMWWt8TPqy

后面几周的嘉宾分享我会边看边拆，关注我，第一时间同步最佳实践🫡

[Mihail Eric @mihail_eric]

I’m excited to finally announce the newest edition my Stanford course 𝗧𝗵𝗲 𝗠𝗼𝗱𝗲𝗿𝗻 𝗦𝗼𝗳𝘁𝘄𝗮𝗿𝗲 𝗗𝗲𝘃𝗲𝗹𝗼𝗽𝗲𝗿. It has been 9 months in the making.

Last November, with the release of Claude Opus 4.5, coding agents experienced a step function improvement in capability. We all felt it. The LLMs were more powerful, could reason for longer, solve harder tasks.

This year’s iteration of my course reflects the 2026 metamorphosis of software engineering.

My core belief is simple: AI-native developers of the LLM era are going to become the most important members of any software organization. I have designed my course to train this next generation of engineers.

𝗪𝗵𝗮𝘁’𝘀 𝗱𝗶𝗳𝗳𝗲𝗿𝗲𝗻𝘁 𝘁𝗵𝗶𝘀 𝘁𝗶𝗺𝗲 𝗮𝗿𝗼𝘂𝗻𝗱

First, 85% of my Fall 2025 class material is being thrown out. The Fall 2026 syllabus reflects the core capabilities AI-native engineers must have: agent skills, advanced context engineering, MCP portals, agent-ready codebase principles, agentic code review, security, parallelizing background agents, software factories, and more.

Second, I am going to teach my students how to have software taste. Every student will be required to ship pull requests to production-grade, real-world codebases. The course is collaborating with the top open-source AI repos who will offer support and mentorship to students on how to meaningfully contribute to their projects.

This has never been done before in any university course so I am incredibly grateful to our OSS Partners: @browserbase, @HeyGen, @CopilotKit, @semgrep, @OpenHandsDev, @milvusio, @marimo_io, Pi, @crewAIInc, @warpdotdev, @vercel, @cmux, @arizeai, @UnslothAI, and @anyscalecompute.

𝗪𝗵𝗮𝘁’𝘀 𝘀𝘁𝗮𝘆𝗶𝗻𝗴 𝘁𝗵𝗲 𝘀𝗮𝗺𝗲

I’m fortunate to again have AI software engineering leaders and founders as guest speakers to share their learnings from building top coding agent products. Thank you to @leerob from @cursor_ai, @bcherny of @claudeai code, @EnoReyes of @FactoryAI, @silasalberti of @cognition, @0xine of @semgrep, Rajesh Bhatia of @Cloudflare , @amasad of @Replit, and @eladgil.

All resources will be available online. All classes will be available to the public.

9/22 on Stanford campus. See you in class. 

https://t.co/wTokHyUMsz

[John Loeber 🎢 @johnloeber]

New Essay

I spent many weekends this year on automated theorem proving with AI + Lean, only for the Bitter Lesson to reign supreme. Some reflections on the OpenAI Math Repo and the short, one-summer window of human-AI collaboration in mathematics.

https://t.co/t23Vd2nhXW

[Anthropic @AnthropicAI]

New on our Frontier Red Team blog: using Claude to formally verify software at large scale.

Formal verification can prove that code is free of certain bugs, but historically it has only been feasible for small codebases (or required a lot of expert time).

With AI models, however, this is beginning to change. In collaboration with @theoremlabs, we used Claude to prove security properties in over one hundred real-world open-source libraries over the span of a few weeks.

Read our research in full here: https://t.co/zUyx8CjBEk

[Former Congresswoman Marjorie Taylor Greene🇺🇸 @FmrRepMTG]

On Lindsey Graham.

Stay with me on this.

The worst kept secret in Washington DC came out today because 3 brave men came out and told the truth and it doesn’t matter if it is after he’s dead.

Brave?

Yes brave.

Why are they brave?

They had to go on camera and tell intimate details of their personal lives.

To the world.

That is extremely difficult to do. Yes it’s brave. And Stressful. They walked into the public square opening themselves to the opinions of the entire world.

No matter your opinion on the issue, they took a courageous leap in coming forward.

If you are critical, would you go on camera and tell your secrets to the entire world?

I would bet a lot of money you wouldn’t.

Ok but there is more to this.

Because this is the world I saw up close in Washington.

I saw behind the curtain and saw ALL the wizards.

Everyone in the highest levels of power in Washington DC are keeping secrets that would destroy their entire lives.

And that’s exactly why they have power.

Not because they are so smart or qualified or the best choice for leadership.

I mean really, if you think our best and brightest are in charge then do some critical thinking and look at the pathetic state of our debt, our dollar, and how our government is run, and the outcome that affects your life.

Their secrets are the only reason these people rise to the top.

Because their secrets are not secrets.

We all knew Lindsey Graham was gay. Everyone did.

Their secrets are levers of control. Keeps them in line. Makes them a team player. And will make them fight the hardest on issues that don’t make sense in order to keep their secrets just that. Secrets.

So everyone is arguing about how this is old news, why now? Or arguing about being gay, and gay issues. Or arguing because that’s what everyone does on the internet, just argue with each other and with bots about issues they don’t have the real truth on, but throw out their strongest opinion anyways.

I’m here to tell you the granular issue you need to dial in on and block out the rest of the noise.

Three gay men just bravely pulled back the curtain and tried to teach you all the most important lesson America needs to learn.

It’s this.

America’s most powerful leaders keep secrets that comprise them in the most dangerous ways for the security, safety, and prosperity of our nation.

That’s the whole point.

I tried my hardest to fight the entire system on both sides of the aisle and couldn’t beat it because YOU ALL keep voting these people back into office and back into power.

By the way, I always cosponsored the Term Limit bill and no matter who is in charge, Pelosi, McCarthy, or Johnson, Dem or Rep, that bill never ever got even close to getting a vote on the House floor.

So the question for you, The American People, is are you able to learn the lesson and if so, what exactly are YOU going to do about it?

Ask yourself, what are you voting for? Donating to? Volunteering for?

And are you really satisfied with the outcome election cycle after election cycle?

[Carole Cadwalladr @carolecadwalla]

The next time Nick Clegg in invited onto BBC to pontificate as an independent tech expert…please note that here he is, named in the $27bn New Mexico judgment as one of execs engaged in a systematic cover-up of Facebook’s deceptive consumer practices.

[Objectively Random @ObjRandom]

@izakaminska Returns are returns to wealth. The stock market may just be incidental. Hugh is thinking in western DM terms.