kdavis

Trephination! :-)

[1] https://en.wikipedia.org/wiki/Trepanning

Surgery, antimicrobials, farming crops, animal husbandry... humans are late to the game.

Using gemma3:12b I ran it once over Andrew Ng's CV https://ai.stanford.edu/~ang/curriculum-vitae.pdf because why not.

He's a 48.0/100, things that make you go Hmmm.

Hmm...six runs with gemma3:12b on my CV

- Varies from 102.0/100 to 100.0/100

- Missed lots of OSS work

- Misinterprets GSoC work (Thinks projects I started that were contributed to in GSoC implies that I received a GSoC stipend)

- Areas for improvement seem to vary inconsistently (There's not enough project detail to there's too much project detail)

I still don't make company cut offs ¯\_(ツ)_/¯

What other verification targets did you consider?

Surprised pikachu face

Time to switch

Ye olde building 20, how we loved to hate you.

The conflict in the Gulf has now claimed several new victims: data centers. This marks a sea change in warfare and will force tech companies to reevaluate their posture around national defense. Defending them effectively means a new policy by the U.S. government—one that creates deterrence not just to protect life and health, but also data.

Almost anything, i.e. the next generation accelerator[1] at CERN is about 15B CHF which is about 20B USD.

[1] https://home.cern/science/accelerators/future-circular-colli...

The 3rd section of the 14th amendment[1] states that no person having engaged in insurrection[2] shall hold any office, civil or military, in the United States. So technically Trump isn’t a legitimate head of state either.

[1]https://en.wikipedia.org/wiki/Fourteenth_Amendment_to_the_Un...

[2]https://www.npr.org/2025/12/31/g-s1-104190/capitol-riot-trum...

10x when working on a code base I'm very familiar with.

Basically, it amounts to being able to give detailed instructions to a junior dev (who can type incredibly fast) and having them carry out your instructions.

If you don't know the code base, and thus can't provide detailed instructions, this junior dev can (using their incredible typing speed) quickly run off the rails. In this case, as you don't know the code base, you wouldn't know it's off the rails. So you're S.O.L.

A simple question about a spinning needle has haunted mathematicians for more than a century. It led to the Kakeya conjecture, a cornerstone of modern analysis connecting geometry, fractals, and the behavior of waves. Now, mathematicians Hong Wang and Joshua Zahl have cracked the 3D case — a once-in-a-generation breakthrough that could reshape how we understand the Fourier transform. (Also featuring Terence Tao and Jonathan Hickman.)

This paper introduces Group Sequence Policy Optimization (GSPO), our stable, efficient, and performant reinforcement learning algorithm for training large language models. Unlike previous algorithms that adopt token-level importance ratios, GSPO defines the importance ratio based on sequence likelihood and performs sequence-level clipping, rewarding, and optimization. We demonstrate that GSPO achieves superior training efficiency and performance compared to the GRPO algorithm, notably stabilizes Mixture-of-Experts (MoE) RL training, and has the potential for simplifying the design of RL infrastructure. These merits of GSPO have contributed to the remarkable improvements in the latest Qwen3 models.

Despite a significant number of up votes in a short amount of time (174 in 1 hour as of this comment a comment to time ratio higher than anything on the front page) this got moved off the front page. Suspicious.

Guth may never get a Nobel.

“The law that entropy always increases holds, I think, the supreme position among the laws of Nature. If someone points out to you that your pet theory of the universe is in disagreement with Maxwell's equations - then so much the worse for Maxwell's equations. If it is found to be contradicted by observation - well, these experimentalists do bungle things sometimes. But if your theory is found to be against the Second Law of Thermodynamics I can give you no hope; there is nothing for it to collapse in deepest humiliation.” ― Arthur Eddington, New Pathways in Science

Given the past week, I think “there's no real possibility of the transatlantic defence alliance suddenly completely failing” is wishful thinking.

In his last term Trump wanted to pullout of NATO[1], but “cooler heads” prevailed. Now these moderating forces are no longer members of his cabinet. So he will doubtlessly try officially, through legal means, or unofficially, by weakening NATO to the point where its guarantees are no longer believed, to withdraw.

The only real moderating force preventing official withdrawal is the “National Defense Authorization Act for Fiscal Year 2024”[2] which “prohibits the President from unilaterally withdrawing from NATO without approval of a two-third Senate super-majority or an act of Congress”.

There is no real moderating force preventing unofficial withdrawal.

[1] https://foreignpolicy.com/2020/06/23/bolton-book-trump-nato-...

[2] https://en.wikipedia.org/wiki/Withdrawal_from_NATO#:~:text=M...

https://en.wikipedia.org/wiki/Budapest_Memorandum

As someone who has done lots of work in both fields, this is some BS.

Goldstein?

Generally this is already "the law of the land" in the US via the HiQ Labs v. LinkedIn precedent[1]. (IANAL)

[1] https://en.wikipedia.org/wiki/HiQ_Labs_v._LinkedIn

It's beta

* 4-Manifolds and Kirby Calculus by Andras I. Stipsicz and Robert E. Gompf

* Differential Manifolds by Antoni A. Kosinski

* Introduction to Smooth Manifolds by John M. Lee

I remember seeing Thad walking around MIT with his wearables on and thinking "Hey, that's weird."

Who'd thunk it; we're all weird now.

I’m disappointed at the level of comments here. HN you’re better than this.

There is a ELIPhD student version here[1].

Personally, though I’ve not read the papers, I doubt that the description of a quantum eigenstate of QED with charge Q can just be copied over mutatis mutandis to describe a quantum eigenstate of quantum gravity with energy E, the theories are just too different.

[1] https://youtu.be/Jn0hsZg_3sE

Links to papers are at the end of the article

You're free to enter any input sentence you want in the text box.

The input sentence generally should be in the language you selected from the dropdown. For example, if the dropdown has "French" selected you could enter the text "Allons enfants de la Patrie, Le jour de gloire est arrivé!"

Clicking "Submit" then generates a TTS reading of the sentence you input in the language selected from the dropdown.

For fun you can mix and match. In other words, select a language from the drop down and enter text in the text box not in the language selected from the dropdown. (For example, the dropdown could have "French" selected and the sentence could be "O say can you see, by the dawn's early light". This gives interesting results, it sounds as if a native French speaker is speaking English.)

We did not provide such a demo in part to hinder nefarious uses of the technology.