Matt Giles

[John H. Boyer @johnhboyer]

anyway here’s the complete works of plato, greek and english side by side, every word parseable, free forever
https://t.co/m4JnhD3aGz

[Kat ⊷ the Poet Engineer @poetengineer__]

i made a reading interface for spinoza and its commentaries throughout centuries, inspire by talmud:
- scroll to adjust each era's thickness
- hover to discover cross-references between commentators

src code for subscribers ↓

[Matt Giles @dmnshingreturns]

Goodhart’s law

[Joe Weisenthal @TheStalwart]

The FT says that Amazon employees are doing random unnecessary task automations to consume tokens and to show their bosses that they're using AI more https://t.co/wZ204CKi32

[History Calendar @HistoryCalendar]

The Four Corners in Kingston, NY is the only intersection in the United States in which all four structures date before the American Revolution.

The oldest, the Jansen House, was built between 1663 and 1667 when the region was part of the New Netherlands colony.

[Matt Giles @dmnshingreturns]

is it bad for the soul, to mistreat language models?

[Matt Giles @dmnshingreturns]

Audible doesn't even have The King's Two Bodies

[Matt Giles @dmnshingreturns]

@gasnerpants It’s definitely both.

[Matt Giles @dmnshingreturns]

It really is some actual solace that everyone already knows that the Trump ballroom will be demolished with slightly more dynamite than is strictly necessary, on national television.

[Matt Giles @dmnshingreturns]

pumped for the approaching wave of “subtly ai” content

[nice handsome boy @edward861312]

oh you're a yankees fan? like donald trump? like henry kissinger? like jeffrey epstein?

[Matt Giles @dmnshingreturns]

Trilemma (Pick 1):

A) Understand primitives/principles of IAM on AWS
B) Understand primitives/principles of IAM on GCP
C) Understand primitives/principles of IAM on Azure

[Matt Giles @dmnshingreturns]

This attack -- which is not obviously not ongoing -- is not unlike the plot to the King novel/show The Outsider.

Every few weeks a new attack. Each attack linked in some way to the last. Every time people think they solved it. But if they had, the next attack wouldn't happen.

[Andrej Karpathy @karpathy]

Software horror: litellm PyPI supply chain attack.

Simple `pip install litellm` was enough to exfiltrate SSH keys, AWS/GCP/Azure creds, Kubernetes configs, git credentials, env vars (all your API keys), shell history, crypto wallets, SSL private keys, CI/CD secrets, database passwords.

LiteLLM itself has 97 million downloads per month which is already terrible, but much worse, the contagion spreads to any project that depends on litellm. For example, if you did `pip install dspy` (which depended on litellm>=1.64.0), you'd also be pwnd. Same for any other large project that depended on litellm.

Afaict the poisoned version was up for only less than ~1 hour. The attack had a bug which led to its discovery - Callum McMahon was using an MCP plugin inside Cursor that pulled in litellm as a transitive dependency. When litellm 1.82.8 installed, their machine ran out of RAM and crashed. So if the attacker didn't vibe code this attack it could have been undetected for many days or weeks.

Supply chain attacks like this are basically the scariest thing imaginable in modern software. Every time you install any depedency you could be pulling in a poisoned package anywhere deep inside its entire depedency tree. This is especially risky with large projects that might have lots and lots of dependencies. The credentials that do get stolen in each attack can then be used to take over more accounts and compromise more packages.

Classical software engineering would have you believe that dependencies are good (we're building pyramids from bricks), but imo this has to be re-evaluated, and it's why I've been so growingly averse to them, preferring to use LLMs to "yoink" functionality when it's simple enough and possible.

[Matt Giles @dmnshingreturns]

I would buy a coffee table book, or like a desk calendar: one page per prediction market trade by insiders.

[The AI Investor @The_AI_Investor]

Nothing to see here

"Polymarket account called "Magamyman" made $515,000 in a single day betting on last night's U.S. strike on Iran, with the first trade placed 71 minutes before the news broke publicly.

When this person bought in, the market had this at a 17% probability. They turned roughly $87,000 into over half a million dollars overnight.

Reminder that Donald Trump Jr. sits on Polymarket's advisory board and his firm invested double-digit millions into the platform last year."

[Matt Giles @dmnshingreturns]

I would buy a coffee table book, or like a desk calendar: one page per prediction market trade by insiders.

[The AI Investor @The_AI_Investor]

Nothing to see here

"Polymarket account called "Magamyman" made $515,000 in a single day betting on last night's U.S. strike on Iran, with the first trade placed 71 minutes before the news broke publicly.

When this person bought in, the market had this at a 17% probability. They turned roughly $87,000 into over half a million dollars overnight.

Reminder that Donald Trump Jr. sits on Polymarket's advisory board and his firm invested double-digit millions into the platform last year."

[Yoni Appelbaum @YAppelbaum]

Albuquerque turned a wall of its airport bookshop over to its local university press, and goddamn, now I want this in every airport.

[Vintage Maps @vintagemapstore]

The City is Alive: The Population of Manhattan, Hour-by-Hour.

Manhattan's nighttime population is 1.6 million, but around 2:00 PM every workday, the island is home to about 4 million people - if Manhattan were a state, it'd be roughly the 28th most populous in the US every afternoon, ranking just above Oklahoma.

Illustration via Joey Cherdarchuk / Darkhorse Analytics

[Matt Giles @dmnshingreturns]

Of all the ways I might have lived long enough to betray the ideals of my youth, none has been so terrible as when I first told my daughter that, if she would like dessert, first I require that she eat her carrots.

[Matt Giles @dmnshingreturns]

Sovereign is that from which normative ideals emerge.

[Robert Anderson @ProfRobAnderson]

Proud to announce my new article, "Hallucinated Cases Are Good Law," forthcoming in the Princeton Law Review.

[John B. Holbein @JohnHolbein1]

“Among articles stating that data was available upon request, only 17% shared data upon request.”