HN user
The only argument they have here is that they use GB300 GPU's which for some reason should not be available to chinese citizens
Note that Chinese companies are free to rent from GB300 clouds internationally. There are large datacenter hubs in Singapore and Malaysia serving chinese and other customers.
Though there is also reported [1] significant smuggling of Nvidia chips into China as well.
On the one hand, these debts may be off the balance sheet, but institutional investors certainly know about them and can reason about the company's valuation. Retail investors may be caught out slightly more.
But on the other hand, these companies are essentially paying for the service of taking the debt off books (by paying the leasing premium to the SPV partners). I guess I'm wondering what they really gain from doing so, if again sophisticated investors can see through the games?
That's the government's problem
Noticed this as well. And the additional definitions under "explore this word" are all relating to the root word, not the actual form of the answer
Definitely agree, I was very confused to see the latest guess (and with the most narrow letter ranges) at the bottom of the list.
Though, by the time we've replicated a complete industrial hinterland (up to and including a semiconductor supply chain) which the author describes, it seems like generating fuel and engines wouldn't be impossible.
Also to the authors last point (extremely long time scales causing degradation), it seems like we'd want high thrust capabilities regardless. i.e. maybe a small gravity well doesn't gain us anything, since we'd need big engines to get up to speed anyway.
I suppose it depends if you're assuming the probe is a complete factory, just taking in regolith and spitting out new probes, vs if the probe deploys and builds up the factory on the surface of an asteroid.
In the latter case spinning doesn't get you far.
So the question is can they keep the pricing up on the older ones a few years down the line
They don't expect to keep the prices flat over time, and everyone involved will have planned for this. Prices are highest when they're the newest and greatest (part of why it's valuable for neoclouds to be first in line for new models), and drop year by year as newer GPU models can do equivalent work at lower cost.
You can see a pretty cool dataset of this at [1]; H100 prices where $3/hr in 2023, and dropped linear-ish to $1.75/hr by 2025. And also the notable exception that prices are up this year due to shortage.
Unrelated to the accomplishment or proof itself, but it's interesting how much of the prompt, even in this latest-and-greatest model, is spent essentially telling the model to actually solve the problem. Things like "Reject status reports, vague optimism, and claims that an unproved global compatibility statement is 'routine'."
Also a lot prompt spent feeding it strategies, which feel like they should/will eventually be deduced by the model itself, not explicitly stated. That's not to take away from the outcome in any way; rather, it feels sort of like when you would prompt GPT 4, "think through your answer step by step," as a sort of proto-chain of thought.
> approximately 700,000 A100e GPU hours of black-box automated red teaming
Amusing that they use A100e as the reference point to sound impressive. Different ways you could make that conversion, but based on FP4 FLOPs (yes it's disadvantageous to A100, that's the point), that's something like 200hr on a GB300 NVL72 rack.
Not nothing either, but far less astounding sounding than 700k hrs.
If I recall correctly, Gibson had never even used a computer at the time of writing Neuromancer, so that's perhaps not shocking.
Uhm.. Super High Frequency (SHF) and Extremely High Frequency (EHF) would like to have a word.
Though yes in my experience, at those frequencies people stop using ITU designations, and switch to IEEE (S,C,X-band etc).
Logic gates and memory bits have very different fabrication processes (mostly because DRAM is optimized for a high density of big capacitors doing the storage).
You can put some memory on the logic wafer (SRAM) but it's area inefficient, which is wasteful on your expensive N2 wafer. So a dedicated DRAM process is vastly cheaper per bit, even at current elevated prices.
I enjoyed that book. But which part are you referring to?
Someone please give Nemotron a thesaurus. This reminds me of the early days of SEO, where you try to hit 1% keyword density.
The new game is finding a single sentence with the most instances of "safe" or "safety". My current high score is 4..
Interesting writeup. Would be nice to have a couple images of Linear A/B scripts to visualize. Looking on google, they're very daunting!
The water is a clever impedance matching trick. The contrast in density between air and human flesh is high, so the waves all reflect off the surface rather than penetrating and reflecting off the internal structures we care about.
That's why normally you're concerned with really good transducer contact (squeezing out any air) or use a gel to match impedance.
I'm a bit rusty on CT, but I'd guess the resolution is proportional to the total number of transducers in the array (e.g. larger sensing surface equals tighter resolution) since you're basically taking a Fourier transform of the incident wave.
This is really interesting! And perhaps surprisingly doesn't trigger any immediate major technical red flags (as someone who has worked with MRI and phased array beamforming), as many HN HW articles do.
My only criticism from the tech video would be that they spend some time lauding the nanometer deflection sensitivity, which might lead some to believe that's indicative of the image resolution. It's not, and it's somewhat of a distraction -- that's just giving us amplitude information, which is comparatively less important than correlated time/phase across the 100k sensors. They do later on state ~mm resolution, which is still great!
Doppler and motion blur may be an issue (e.g. heart beating), as one slice requires a full ring of sequential exposures. But still way faster than MRI, so probably fine.
On a lighter note, it could seriously change the meaning of get FUCT (Full body Ultrasound Computational Tomography)!
We already ultrasound babies in the womb, so one would hope this has been studied.
To be clear, fuel cells are considered "low air pollution" because they eliminate certain nasty combustion products (NOx), but they still produce as much CO2 per kWh as a gas turbine.
Arguably that CO2 stream is concentrated and a candidate for capture/sequestration, but no one is doing that in practice.
Also gas turbines are on backorder until 2029, and new grid connection queues on a similar timescale.
Hence why all the bitcoin miners are cashing in (or trying to) by converting their facilities to datacenters.
I think the right comparison would be a vertically integrated Neocloud like Oracle (insofar as they own some/most of their own datacenters, unlike a CoreWeave).
Minor point, but is there a trademark issue here with Yahoo Scout? (the Yahoo finance AI summaries widget)
Given the $10k price tag for tokens and high rate of bugs (several per minute) they mention, it'd be very interesting to see this experiment run with cheaper models too.
I wonder if we get to a world where a full repo sweep like this is a default Github action after commit.
The wild thing to me, is that they're serving $47B run rate worth of requests on maybe 2-3 GW of compute currently [1], of which only a fraction goes to inference, vs R&D and training. Obviously there have been complaints on token limits and such so they're stretched a bit thin, but nonetheless.
Hard to imagine what a world with 100GW of compute looks like.
[1] https://epochai.substack.com/p/frontier-labs-dont-use-most-a...
^^ This quotes 1.4GW at the end of 2025. Add 0.3GW at Colossus 1, and some initial fraction of 1GW Trainium2 from [2]
Lately I've been thinking that UI really needs to include the equivalent of a screenshare meeting. Ideally you could click through an example of a software flow Claude's never seen before, with a few quick notes, and have it reliably work.
These narrow integrations with specific software suites seems like a dead end.
I had a similar, really great prof, who would always ask for what the next variable would be, so we'd end up with trees and smiley faces. His point was to not make assumptions (c is always a constant etc), but it made the classes more engaging too.
And, somehow every example ended along the lines of "then you hand this to your boss, kick up your feet and have a nice glass of scotch."
I think the water is difficult to traverse, in that it slows you down when 'swimming'.
It's really interesting how it still feels grounded even though you can fly all around. Having the cursor disappear underneath bridges and behind buildings really helps the illusion.
Do we know the breakdown of revenue from API vs subscriptions for OAI/Anthropic? That seems very relevant, since this entire article seems to be on the premise that users are only willing to pay for a subsidized subscription and would never pay the 'true' token cost.
The internet seems to be saying that 70%+ of Anthropic revenue is per-token metered API, which would largely invalidate the article, but I can't find a solid source.