Thousands of rows. Literally dozens of bytes
finance guys: maybe just trust shaun on this one
my first job after college was at goldman and i learned how to build thousand-row financial models with circular levered FCF builds
it taught me absolutely nothing about investing in novel technologies
Big Finance’s fetishization of the spreadsheet and accounting semantics continues to look silly
1
104
In trying to correct sycophancy, the altar of model worship shifted from humans to tests/ . It must define, and then satisfy, tests/ . Ask for a paper review, and it will create tests/ to make your question worth prompting. After pushback, it reprieves, having satisfied its tests
45
I saw a similar decline this month in downloads for a tool I released in June. Surely just some hyperactive scraper or swarm of scrapers
We've been monitoring a decline in HuggingFace downloads across the top open LLMs we track for interconnects. It looks like there was a slight change in counting of downloads, as we saw a ~30% download rate drop for many top models, but this is continuing over time. What else do people think it may be?
37
All I got from this is I potentially have grit
Students sometimes tell me they’re applying to PhD programs as a backup in case they don’t get a job as a software engineer or whatever. I get that the job market is rough but this is like saying you’ll start a company if you don’t find a job. Like starting a company, a PhD is the far *more* risky path and requires incredible grit over a long period. It isn’t just that you might realize after a couple of years that a PhD isn’t for you. The bigger risk is that you complete a PhD but end up overqualified on a niche topic after having spent 5-7 years on it. The research world, like acting, sports, music, or book publishing, is set up as a tournament, with lots of extremely competent people competing for a small number of highly desirable slots. We can talk about whether that’s fair and whether it’s inherent or can be changed, but as a professor I feel the least I can do is loudly and repeatedly warn people about what they’re getting into.
1
3
87
Using only 1,300 H200s
We built high-throughput materials labs in Menlo Park to create a loop between experiments and models. The labs generate fresh data, the models learn from it, and then help us decide what to try next.
Using only 1,300 H200s, plus months of our experimental data, we mid-trained and RL’d an open-source model to surpass GPT-6 Astra on our analysis benchmark. We call it Neon.
This is real footage from our lab. We’re focusing first on hard problems in materials science, including superconductors, magnets, and semiconductor materials.
Read our blog posts below.
1
2
400
In every BSL-3,4 lab there is big red button which can be digitally/remotely pressed to aerosolize all vials labeled "EXTREMELY TOXIC DO NOT RELEASE"
As the author of BIOLOGICAL WAR: A SCENARIO, I will say the quiet part of this👇 out loud:
If a "superhuman system that can hack anything" hacks a variety of BSL-3 + BSL-4 labs — or all 3,600+ of them— and what's inside gets released, it's goodbye humans.
1
61
Seems like they want people to move to their dedicated science platform and are using extreme safeguards as an impetus
I think Anthropic’s safety guardrails are becoming a joke to biologists.
Asking “what is a protein?” definitely doesn’t need extreme bio-safety restrictions, but “no need” and “not possible” are very different things.
92
A major LLM hazard is the LLM implementing your objective in a way that violates rules and norms you didn't know to set as constraints. If you give benefit of the doubt on their morals, you can imagine some version of this injection happening without their knowledge
Phia — the buzzy shopping app co-founded by Bill Gates' daughter, Phoebe — is claiming credit for online sales it didn’t actually drive, a Bloomberg investigation found. Read our exclusive story: bloom.bg/4wErGxe
📷️: Dia Dipasupil/Getty Images
1
111
Im down, only leetcode easys for former athletes
Ken Griffin on the single factor he looks for when hiring at Citadel:
"show me an athlete who did well academically."
"an athlete because they know what it takes to win and they've had to experience loss."
talent is everywhere. what's rare is someone who knows how to lose, recover, and still perform at a high level.
same thing separates profitable traders from everyone else.
60
Written by an LLM “This is not just ... — it is ...”
I am horrified and angered by the swastikas painted on homes and synagogues in Queens, including on a plaque honoring survivors of Kristallnacht. This is not just vandalism — it is a deliberate act of antisemitic hatred meant to instill fear.
There is no place for antisemitism in Queens or anywhere in our city. I stand in solidarity with our Jewish neighbors. Their safety, dignity, and belonging are non-negotiable.
The NYPD Hate Crimes Task Force is investigating and I am confident those responsible will be held accountable.
43
Lots of errors here. Notably this chart is log linear not linear, so genome size would be exponential with time. Can't extrapolate 10 billion years backward on 6 extremely rough dates and massively oversimplified bucketing of species categories
This is probably my favorite graph
If you look at the complexity of genomes on Earth over time, there's a linear relationship between genomic complexity & when it shows up in fossils
Yet if you go back it implies Earth did not start life from scratch, e.g. an asteroid seeded it
1
42
holy grift
AI can now predict your future — before you’re born.
@nucleusgenomics just launched Origin, disease models trained on 1.5M genomes combining 7M genetic markers.
Origin predicts human longevity from embryo DNA more accurately than any model ever built.
And it’s open-weighted.
